v4(1) 2015:: american transactions on engineering & applied sciences

IN THIS ISSUE

American Transactions on Engineering & Applied Sciences

Volume 4 Issue 1 (January 2015)

ISSN 2229-1652 eISSN 2229-1660

http://TUENGR.COM/ATEAS

http://tuengr.com/ATEAS/V04/001.pdf

















http://tuengr.com/ATEAS



http://TuEngr.com/ATEAS

International Editorial Board Editor-in-Chief Zhong Hu, PhD Associate Professor, South Dakota State University, USA

Executive Editor Boonsap Witchayangkoon, PhD Associate Professor, Thammasat University, THAILAND

Associate Editors: Associate Professor Dr. Ahmad Sanusi Hassan (Universiti Sains Malaysia ) Associate Prof. Dr.Vijay K. Goyal (University of Puerto Rico, Mayaguez) Associate Professor Dr. Narin Watanakul (Thammasat University, Thailand ) Assistant Research Professor Dr.Apichai Tuanyok (Northern Arizona University, USA) Associate Professor Dr. Kurt B. Wurm (New Mexico State University, USA ) Associate Prof. Dr. Jirarat Teeravaraprug (Thammasat University, Thailand) Dr. H. Mustafa Palancıoğlu (Erciyes University, Turkey ) Editorial Research Board Members Professor Dr. Nellore S. Venkataraman (University of Puerto Rico, Mayaguez USA) Professor Dr. Marino Lupi (Università di Pisa, Italy) Professor Dr.Martin Tajmar (Dresden University of Technology, German ) Professor Dr. Gianni Caligiana (University of Bologna, Italy ) Professor Dr. Paolo Bassi ( Universita' di Bologna, Italy ) Associate Prof. Dr. Jale Tezcan (Southern Illinois University Carbondale, USA) Associate Prof. Dr. Burachat Chatveera (Thammasat University, Thailand) Associate Prof. Dr. Pietro Croce (University of Pisa, Italy) Associate Prof. Dr. Iraj H.P. Mamaghani (University of North Dakota, USA) Associate Prof. Dr. Wanchai Pijitrojana (Thammasat University, Thailand) Associate Prof. Dr. Nurak Grisadanurak (Thammasat University, Thailand ) Associate Prof.Dr. Montalee Sasananan (Thammasat University, Thailand ) Associate Prof. Dr. Gabriella Caroti (Università di Pisa, Italy) Associate Prof. Dr. Arti Ahluwalia (Università di Pisa, Italy) Assistant Prof. Dr. Malee Santikunaporn (Thammasat University, Thailand) Assistant Prof. Dr. Xi Lin (Boston University, USA ) Assistant Prof. Dr.Jie Cheng (University of Hawaii at Hilo, USA) Assistant Prof. Dr. Jeremiah Neubert (University of North Dakota, USA) Assistant Prof. Dr. Didem Ozevin (University of Illinois at Chicago, USA) Assistant Prof. Dr. Deepak Gupta (Southeast Missouri State University, USA) Assistant Prof. Dr. Xingmao (Samuel) Ma (Southern Illinois University Carbondale, USA) Assistant Prof. Dr. Aree Taylor (Thammasat University, Thailand) Assistant.Prof. Dr.Wuthichai Wongthatsanekorn (Thammasat University, Thailand ) Assistant Prof. Dr. Rasim Guldiken (University of South Florida, USA) Assistant Prof. Dr. Jaruek Teerawong (Khon Kaen University, Thailand) Assistant Prof. Dr. Luis A Montejo Valencia (University of Puerto Rico at Mayaguez) Assistant Prof. Dr. Ying Deng (University of South Dakota, USA) Assistant Prof. Dr. Apiwat Muttamara (Thammasat University, Thailand) Assistant Prof. Dr. Yang Deng (Montclair State University USA) Assistant Prof. Dr. Polacco Giovanni (Università di PISA, Italy) Dr. Monchai Pruekwilailert (Thammasat University, Thailand ) Dr. Piya Techateerawat (Thammasat University, Thailand ) Scientific and Technical Committee & Editorial Review Board on Engineering and Applied Sciences Dr. Yong Li (Research Associate, University of Missouri-Kansas City, USA) Dr. Ali H. Al-Jameel (University of Mosul, IRAQ) Dr. MENG GUO (Research Scientist, University of Michigan, Ann Arbor) Dr. Mohammad Hadi Dehghani Tafti (Tehran University of Medical Sciences)

2015 American Transactions on Engineering & Applied Sciences.


Contact & Office:

Associate Professor Dr. Zhong Hu (Editor-in-Chief), CEH 222, Box 2219 Mechanical Engineering

Department, College of Engineering, Center for Accelerated Applications at the Nanoscale and Photo-Activated Nanostructured Systems, South Dakota Materials Evaluation and Testing Laboratory (METLab), South Dakota State University, Brookings, SD 57007 Tel: 1-(605) 688-4817 Fax: 1-(605) 688-5878

[email protected], [email protected] Postal Paid in USA.


ISSN 2229-1652 eISSN 2229-1660 http://tuengr.com/ATEAS

Original art version :: heart vine3 clip art available from http://www.clker.com/clipart-heart-vine3.html

FEATURE PEER-REVIEWED ARTICLES for Vol.4 No.1 (January 2015)

1

13

31

39

49

mailto:[email protected]

mailto:[email protected]


















:: American Transactions on Engineering & Applied Sciences


Call-for-Papers:

ATEAS invites you to submit high quality papers for full peer-review and possible publication in areas pertaining to our scope including engineering, science, management and technology, especially interdisciplinary/cross-disciplinary/multidisciplinary subjects.

Next article continue



*Corresponding author (Sudsaisin Kaewrueng). Tel/Fax: +66-2-561-3482. E-mail address: [email protected]. 2015. American Transactions on Engineering & Applied Sciences. Volume 4 No.1 ISSN 2229-1652 eISSN 2229-1660 Online Available at http://TUENGR.COM/ATEAS/V04/001.pdf .

1



A Suitability Comparison among Four Hydroponic Solutions for Growing Lettuce (Lectuca sataiva L. var. green oak)

Radika S. Malawwathanthri a, Sudsaisin Kaewrueng

b*,

Somchai Anusontpornperm c

, and Thunya Taychasinpitak d

a Sustainable Agriculture Program, Faculty of Agriculture, Kasetsart University, Bangkok

10900, THAILAND. b Department of Farm Mechanics, Faculty of Agriculture, Kasetsart University, Bangkok 10900,

THAILAND. c Department of Soil Science, Faculty of Agriculture, Kasetsart University, Bangkok 10900,

THAILAND. d Department of Horticulture, Faculty of Agriculture, Kasetsart University, Bangkok 10900,

THAILAND. A R T I C L E I N F O

A B S T R A C T

Article history: Received 29 August 2014 Received in revised form 02 October 2014 Accepted 08 October 2014 Available online 09 October 2014

Keywords: Nutrient solution; Crop content; NFT system; ANOVA; LSD.

This study compared the suitability of four hydroponic solutions, solution y commonly used in Sri Lanka, solution x vastly used in Thailand and solution z1 and z2 prepared using solution y, for growing lettuce using Nutrient Film Technique (NFT) system. Due to a lower amount of N in solution y than in solution x, the adjustment of this nutrient in the former solution to half and equal amounts of the latter solution, was done using Ca(NO3)2 and KNO3. During a growing period, all solutions pH was maintained to 6.0-6.5. EC level of the solutions and number of leaves were weekly monitored. Fresh and dry upper ground biomasses were weighed and plant tissue samples were analyzed for the concentration of N, P, K, Ca and Mg. Analysis of variance (ANOVA in SPSS/FW) as completely random design and means among treatments were compared using least significant difference (LSD) and different being tested at 0.05 probability. Result showed that although dry matter yield is similar, growth and development of lettuce was better when grown in solution x than in other three treatments. N and K content in solution x and K content in solution y should be reduced. On the other hand, P content in both solutions should be increased to raise P content in the upper ground biomass up to normal range.

2015 Am. Trans. Eng. Appl. Sci.


2 R. S. Malawwathanthri, S. Kaewrueng, S. Anusontpornperm, and T. Taychasinpitak

1. Introduction Hydroponics is a soilless culture used for green house farming. Nutrient solutions are used to

supply nutrients for crop so that physiological requirements of plants can be met without the use of

soil. Hydroponic cultivar allows growers easily to control the nutrient supply, by adjusting the

concentration of the hydroponic nutrient solution. This factor affects the plant water and salt

relations and influences plant growth and quality (Caruso et al., 2011). The plants can grow in an

ideal environment, since everything can be determined, which is normally up to “mother nature”.

In a completely controlled environmental agricultural system, light, temperature, water, carbon

dioxide, oxygen, pH, Relative Humidity (RH), and nutrients are controlled (Douglas, 1975).

In this experiment, hydroponics culture was used because vegetables can be grown under

hydroponics in intensive agriculture. Nowadays, intensive agriculture is very important to

promote because of technology development. There are so many technologies to use in

agriculture using hydroponics. Hence development of cultivation can be improved to use those

methods and plants can protect from adverse and unexpected weather conditions such as heavy

rain, strong winds, hailstorms, frost, snow, hail, low temperatures, high heat and excessive

sunlight. They are also protected from pests and plant diseases, thus avoiding pesticides and

herbicide residues that contaminate.

At this time and also in the near future, arable lands for cultivation will be limited because of

rapid growth rate of population. Hydroponic is an alternative of which growers can produce their

own food on their balcony or other free spaces within the area of their house using pots, bottles,

cups etc., that are available at home. Besides, environmental disasters can happen easily, resulting

in food shortage; therefore hydroponics can become very essential to people, living in those areas.

However, hydroponic techniques still need to be improved in order to secure a certain outcome.

Hence, this research was conducted to investigate a suitable hydroponic solution that can be used

for growing lettuce by using Nutrient Film Technique (NFT) system.

2. Materials and Methods Two commercial hydroponic solutions and two self-made solutions which were made using

former one solution according to N level of other solution were used. Completely randomized

design (CRD) with four replications was employed in this study. Lettuce (var. green oak)

seedlings were grown in four different hydroponic solutions as follow: solution x (Treatment 1;


3

T1), solution y (Treatment 2; T2), solution y + adjacent N to equivalent N existed in solution x;

solution z1 (Treatment 3; T3) and solution y + half of adjacent N added in solution z1; solution z2

(Treatment 4; T4).

At the beginning two NFT continuous circulating systems were installed. One unit

contained four tunnels with 3 m length. Each tunnel had 14 holes with 20 cm spacing. Tunnel

bed was set 1:22 (4.5%) slope and its systematic flow rate was 3 L/minute. After that solutions

were prepared as follows

2.1 Solution x (T1) Stock solution was firstly prepared using A, B and C packets with proportional content of

nutrients. A consist of CaNO3, B consist of KH2PO4 and MgSO4 and C consist of other nutrient

elements. Nutrients were arranged as three packets because of preventing nutrient precipitation in

the solution. A, B & C were dissolved in three separate 5 liters cans, water was poured into cans

and nutrients were added into cans as specified. Every can was shaken well and volume of each

stock solution was adjusted to 5 liters A 1,000-liter solution can be made from these stock

solutions. In this experiment 40 liters of growing solution was prepared using 200 ml of stock

solutions from each can.

2.2 Solution y (T 2) Powdered nutrient pack (1000 grams) was dissolved in 500-liter water. Hence in this

experiment 80 g of nutrient powder was dissolved in 40 liters of water.

2.3 Solution z1 (T3) In this treatment, solution y was modified by adding adjacent N in the form of Ca(NO3)2 and

KNO3. The amount added leads to raising N equivalent to N existed in solution x.

2.4 Solution z2 (T4) This treatment was similar to T3 but with the half amount of adjacent N used in that treatment.

2.5 Seed Planting Seeds of lettuce (green oak) were planted in standard hydroponic cups for NFT system which

filled with seed germination media, consisting of perlite and vermiculite in the ratio of 1:4 by

volume. After the media was soaked with water, seeds were placed into a hole in the media to the


depth of 0.5 to 1cm at the rate of one seed per cup. In the first 5th days, tanks were filled with

water and a pump was operated, greenhouse roof was covered by a shading sheet for seed

germination. After the 5th day of the seedlings stage, water in tanks was replaced by the respective

solutions and shading cover of green house was removed.

The pH of the solution was steadily maintained at 6.0-6.5 while electrical conductivity (EC)

was not maintained but the level was recorded. Water level in the tank was daily controlled (40

litters). In this experiment hope to evaluate the conditions of four treatments without maintaining

EC level because effect of the EC should be evaluated. After the 2nd week, 4th week, 5th week and

between 5th and 6th week 10, 5, 5 and 15 liters of solutions was added respectively to every

treatment because EC level of solution y (T2) showed a less nutrient amount that nearly to

minimum level. After 40 days of the crop, plants were harvested.

2.6 Data Recording Following measurements were undertaken to assess a performance of plant growth in different

hydroponic solutions, pH and EC values were measured weekly, using multi-parameter PCTestrTM

35 instrument of which solution temperature can also be measured. Number of leaves was counted

weekly, starting from first week planting to harvesting time, N, P, K, Ca and Mg concentrations in

each solution were analyzed before planting and after harvesting, Fresh weight and dry weight of

the each lettuce were measured after harvesting, N, P, K, Ca and Mg concentrations in each plants

were also analyzed.

2.7 Data Analyzing Nitrogen was determined by Kjeldhal method (Gallaher et al., 1976, Jackson, 1965), P was

determined by Vanadomolibdate method (Johnson and Ulrich, 1959; Westerman, 1990) and K, Ca,

Mg were determined by Atomic Absorption Spectrophotometry method (AAS) (Johnson and

Ulrich, 1959; Westerman, 1990). Analysis of variance (ANOVA) was performed by using

SPSS/FW (Statistical Package for Social Science for Window) for data analyzing as completely

random design and means among treatments were compared using least significant difference

(LSD) and different being tested at 0.05 probability (p<0.05).

3. Results and Discussion

3.1 Electrical Conductivity of Hydroponics Solutions Electric conductivity of four solutions during a period of conducting experiment is shown


5

in Figure 1 and nutrient content in each solution before planting lettuce is shown in Figure 2.

EC of hydroponic solutions varied among four treatments during cropping cycle with values

ranging between 0.8-2.5 mS.cm-1. This result indicates that the fluctuation of EC in hydroponic

solutions from this study was in the range (0.5 to 4.5 mS.cm-1) for normal plant growth (Schwarz,

1995; Jones, 1997). Cooper, 1976 said that in earlier studies, investigators and growers

measured the electrical conductivity of the solution and used the results as an intuitive indication

of the total amount of nutrients and salts. As EC level of T1 is higher than other three treatments

and EC level of treatments can be arranged as T1>T3>T4>T2, it can be said that nutrient content

in each treatment can be arranged as T1>T3>T4>T2. Nutrient content in each solutions before

planting can be arranged as T1>T3>T4>T2 (Figure 2) and hence EC in T1 showed high level in

T1 than other three treatments (T1>T3>T>T2).

Figure 1: Changes of EC values in four different hydroponic solutions in each week.

3.2 Plant Growth and Development

3.2.1 Number of leaves per plant

Average number of leaves per plants in four hydroponic solutions is illustrated in Table 1 and

Figure 3. There was not significant difference in leaf number among treatments in the first three

weeks of measurement. But significant differences after planting from the 4th week until

harvesting time.

0.0

0.5

1.0

1.5

2.0

2.5

3.0

1 2 3 4 5

EC

val

ue (

mS

.cm

-1)

Week

T1 T2 T3 T4


Figure 2: Nutrient content in each hydroponic solutions before planting lettuce.

Table 1: Average number of lettuce leaves from four hydroponic solutions

after planting in each week Treatment 1st week 2nd week 3rd week 4th week 5th week 6th week

T1 T2 T3 T4

1a 1a 1a 1a

3a 3a 3a 3a

5.95a 5.90a 5.75a 5.85a

10.95a 9.90b 9.25c 9.10c

12.95b 12.90b 13.80a 12.85ab

24.10a 21.90c 22.65bc 23.65ab

a, b, c = in the same column was not Significantly different ( p < 0.05).

Figure 3: Average number of lettuce leaves per plant in each hydroponic solutions

after planting from each week.

245.

0

69.3

255.

6

278.

8

94.4

175.

0

72.4

263.

6

205.

9

20.3

245.

0

72.1

294.

6

278.

9

19.5

201.

0

72.6

277.

3

238.

8

19.7

0

50

100

150

200

250

300

350

N P K Ca Mg

Nut

rien

t con

tent

bef

ore

plan

ting

(p

pm)

Nutrient

T1

T2

T3

T4

0

5

10

15

20

25

30

1 2 3 4 5 6

Num

ber

of le

aves

/pla

nt

Week

T1

T2

T3

T4


7

According to the analysis it can be said that lettuces grown in solution x (T1) is better in 4th

5th and 6th week than in other three treatments (T2, T3 and T4) may be due to treatment T1 has

more N content than in T2 and T4 (Figure 2) and more N uptake than in other three treatments

(Figure 6 and Figure 8). Water uptake rate is higher in T1 than in T3 and T4 (Figure 5). Therefore N

uptake is also high in T1 because NO3- is absorbed by passive absorption (Jones, 1997).

3.2.2 Fresh yield and dry matter yield of upper ground biomass of the lettuce

Fresh yield and dry matter yield of the upper ground biomass of lettuce harvested grown in

different hydroponic solutions experiment are shown in Table 2. Fresh yield of lettuce was

statistically different among treatments whereas there was no difference in dry matter yield.

Table 2: The effect of four hydroponic solutions on fresh and dry matter yield of lettuce

Treatment Fresh yield (g/plant) Dry matter yield (g/plant) T1 T2 T3 T4

88.199 a 67.815c

73.195 bc

74.394 bc

2.723 a 2.489 a 2.241 a 2.391 a


Fresh yield from upper ground biomass of the lettuce in T1 showed significant difference

from T2, T3 and T4. But there were no significant differences between T2 and T3, T2 and T4, T3

and T4.

As shown in Figure 5, water uptake is the highest in T2 but the number of leaves per plant in

harvesting stage (Figure 3) is the lowest because of low nutrient content in the solution. Hence,

fresh yield in this treatment is also lower than those of the other three treatments. Although N

content in T3 was greater than in T4, fresh yield is low in T3 than in T4 because of high water

uptake (Figure 5) and number of leaves at the harvesting stage (Figure3) was higher in T4 than in

T3.

Considering above results it can be concluded that lettuces grown in T1 is better than the

other treatments. According to number of leaves and fresh yield from the upper ground lettuce

biomass, it can be concluded that plant solution x (T1) is suitable for lettuce growth and

development than the other treatments in NFT system.


Figure 5: Average water uptake from each lettuce which grew in different hydroponic

solutions in each week..

3.3 Nutrient Content in the Upper Ground Biomass of the Plant N, K, P, Ca and Mg concentrations in the upper ground biomass of lettuce grown in different

hydroponic solutions are shown in Figure 6. There were significant difference in N, P, K, and Ca

in four treatments but Mg content were no significant difference.

Table 3: The effect of four hydroponic solutions on nutrient content in upper ground Biomass of lettuce

Treatment N content g.kg-1

P content g.kg-1

K content g.kg-1

Ca content g.kg-1

Mg content g.kg-1

T1 T2 T3 T4

49.875a 44.822b

47.097ab 45.391b

2.017 c 3.276b 3.626a 3.622a

107.976a 94.196c

108.176ab 106.691ab

18.809 c 26.355a

23.853ab 22.842b

7.864a 8.450 a 7.034 a 6.990 a


It is evident that N content in T1 (solution x) and T3 were higher than in other two treatments

( T2 and T4) as showed in Table 3. According to the analysis, it can be concluded that P content in

upper ground biomass of the lettuce from T3 and T4 are better than those in T2 and T1. K content

in T1, T3 and T4 are better than in T2. Ca content in T2 and T3 are better than those in T1 and T4.

From the result of Jones, 1998 reported that sufficiency range for nutrient in aerial part of

lettuce is N- 35 to 45 g.kg-1, P - 4 to 8 g.kg-1, K-55 to 62 g.kg-1, Ca- 20 to 28 g.kg-1 and Mg - 6 to 8

g.kg-1. According to Table 3, N content of upper ground biomass of the lettuce in T2 is in

0

10

20

30

40

50

60

70

1 2 3 4 5 6

Wat

er u

ptak

e (l

iter

/ w

eek)

Week

T1

T2

T3

T4


9

sufficiency range and other three treatments are greater than in sufficiency range. Jones, 1998 said

that NO3 may begin to accumulate in the plant to fairly high concentrations if there is a substantial

N supply in the rooting media. Hence it can be said that as N content is higher in T1, T3 and T4 than

in T2 (Figure 2), N accumulation also high in these three treatments.

K content in four treatments is greater than in sufficiency range. Jones, 1998 reported that most

plants absorb more K than they need and it’s frequently referred to as luxury consumption. P

content of upper ground biomass of the lettuce in four treatments is lower than in sufficiency range.

Ca content in T2 is higher than in sufficiency range and other three treatments are in sufficiency

range. Mg content in T2 is greater than in sufficiency range and but other three treatments are in

sufficiency range. Therefore it can be concluded that K content and N content in T1, T3 and T4

should be reduced while reducing K content in T2. P content in all treatments should be increased.

Figure 6: Average nutrient content from the upper ground biomass of lettuce in four treatments.

3.4 Nutrient Content and Nutrient uptake of each Treatment Nutrient content in each hydroponic solution before planting is shown in Figure 2, total added

nutrient in each treatment is shown in Figure 7 and nutrient in residue after harvesting lettuce from

each treatment is shown in Figure 8.

0

20

40

60

80

100

120

N P K Ca Mg

Nut

rien

t con

tent

(g.

kg-1

)

Nutrient

T1

T2

T3

T4


Figure 7: Total added nutrient concentration in four hydroponic solutions.

Figure 8: Nutrient concentrations residues in four hydroponic solutions

after harvesting lettuce already.

When considering nutrient content before planting plants and total added solution, it can be

said that N content in T1 was similar to T3 and T1= T3>T4>T2. P content is in four treatment were

nearly the same as K but K content was T3>T4>T2>T1. Ca content was T1=T3>T4>T2. Mg

content in T1 was the highest and in T2, T3 and T4 nearly same content as shown in Figure 2 and

Figure 7.

In Figure 8, the residue content of N and K after harvesting is the lowest in treatment T1 but P,

Ca and Mg remains the lowest in treatment T2. According to EC of the solution, it shows that EC of

459.

4

130.

0

479.

2

447.

5

133.

1

328.

1

135.

7

494.

2

364.

6

21.2

459.

4

135.

3

552.

3

477.

1

14.1

376.

9

136.

1

519.

9

412.

7

14.1

0

100

200

300

400

500

600

N P K Ca Mg

Tota

l add

ed n

utri

ent (

ppm

)

Nutrient

T1

T2

T3

T4

020406080

100120140160180

N P K Ca Mg

Nut

rien

t res

idue

(pp

m)

Nutrient

T1

T2

T3

T4


11

the treatment T2 was the lowest. Hence nutrient content of the residue is also low in treatment T2.

As treatment T1 is shown a high growth rate than in treatment T2, it can be said that N was more

uptake by the treatment T1 than in treatment T2. Hence, treatment T1 residue was low in N content.

4. Conclusion Lettuce growth and development in solution x (T1) is better than in solution y (T2), solution z1

or solution y adjacent N to equivalent N existed in solution x; by adding calculated amount of

Ca(NO3)2 and KNO3 and solution z2 or solution y supplemented with half the amount of Ca(NO3)2

and KNO3 which used in z1 in NFT system.

According to nutrient content in the upper ground biomass of the lettuce, N and K content in

solution x and K content in solution y should be reduced up to suitable level which can be obtained

N and K in upper ground biomass at the usual range of 35 to 45 g.kg-1 and 55 to 62 g.kg-1

respectively. While P content in both solutions should be increased up to suitable level for raising P

in upper ground biomass to the optimum level of 4 to 8 g.kg-1.

5. Acknowledgements This research was financially supported by Kasetsart University and Thailand International

Co-operation Agency (TICA). Dr. C.Yongyut was gratefully acknowledged.

6. References Caruso, G., G. Villari, G. Meichionna and S. Conti. 2011. Effect of cultural cycles and nutrient

solution on plant growth, yield and fruit quality of alpine strawberry (Frageria vesca L.) grown in hydroponics. Scientia Horticulturae. 129: 479-485.

Cooper, A.J. 1976. Crop production with nutrient film technique. IWOSC Symp., Las Palmas. pp 121-136.

Douglas, J.S. 1975. Hydroponics, 5th edition. Oxford Press Bombay. pp 1-5.

Ellis, C. and M.W. Swaney. 1938. Soilless Growth of Plants. Reinhold Publishing Corporation, New York. pp 30 – 44.

Gallaher, R.N., C.O. Weldon and F.C. Boswell. 1976. A semiautomated procedure for total N in plant and soil samples. Soil Science Society Of America. 40: 887-889.

Jackson, M.L. 1965. Soil Chemical Analysis Advanced Course. Department of soils, University


of Wisconsin, USA.

Jones, J. B. 1998. The Essential Elements: Plant Nutrition. CRC Press, LLC. pp 5-12.

Jones, J.B. 1997. Systems of Hydroponics Culture: Hydroponics A practical Guide for the Soilless Grower. St Lucie Press, Florida. pp 106.

Johnson, C.M. and A.Ulrich. 1959. Analytical methods for use in plant analysis.

California Agricultural Experiment Station Bulletin. 767: 25-78

Schwarz, M. 1995. Soilless Culture Management. Spring-Verlag, Germen. pp 3-16.

University of Minnesota Extension. 2009. Growing Salad Vegetable Crops http://www.extension.umn.edu/distribution/horticulture/DG0434.html, August 14, 2012

Westerman, R.L. 1990. Soil Testing and Plant Analysis, 3rd ed. American Society of Agronomy and Soil Science Society of America, Madison, Wisconsin.

Winterborne, J. 2005. Hydroponics: Indoor Horticulture. Pukka Press. pp 113.

Radika Samanthi MALAWWATHANTHRI is a master degree student in Faculty of Agriculture, Kasetsart University, Bangkhen, Bangkok, THAILAND. Her research encompasses hydroponics.

Dr.Sudsaisin KAEWRUENG is an Assistant Professor in Department of Farm Mechanics, Faculty of Agriculture, Kasetsart University, Bangkhen, Bangkok, THAILAND. He holds a Ph.D.(Integrated Water Resources Management) from Asian Institute of Technology, Thailand. He is teaching and researching in farm mechanization.

Dr.Somchai ANUSONTPORNPERM is an Assistant Professor and Head of the Department of Soil Science, Faculty of Agriculture, Kasetsart University, Bangkhen, Bangkok, THAILAND. He obtained his bachelor and master degrees from Kasetsart University. Later, he earned a PhD (Soil Science) from The University of Reading, UK. He is teaching and researching in soil science for agriculture.

Thunya TAYCHASINPITAK is an Associate Professor in Department of Horticulture, Faculty of Agriculture, Kasetsart University, Bangkhen, Bangkok, THAILAND. He is teaching and researching in floriculture and floriculture crop improvement.

*Corresponding author (T.Higashimachi). Tel/Fax: +81-96-326-3479. E-mail address: [email protected]. 2015. American Transactions on Engineering & Applied Sciences. Volume 4 No.1 ISSN 2229-1652 eISSN 2229-1660 Online Available at http://TUENGR.COM/ATEAS/V04/013.pdf .

13



Finite Element Analysis of the Human Middle Ear and an Application for Clinics for Tympanoplasty (Static and Harmonic Vibration Analysis)

Takao HIGASHIMACHI a*

, and Ryuzo TORIYA b

a Department of Mechanical Engineering, Faculty of Engineering, Sojo University, JAPAN. b Toriya ENT Clinic, JAPAN.

A R T I C L E I N F O

A B S T R A C TArticle history: Received 30 September 2014 Received in revised form October 20, 2014 Accepted October 27, 2014 Available online 30 October 2014 Keywords: Geometric model; FEM; Vibration analysis; Human middle ear; Auditory ossicles; Hearing ability; Columella.

We have already proposed a method for estimating the hearing restoration effect of the tympanoplasty operation using three-dimensional finite element static analysis (Higashimachi et al., 2013). In this study, the restoration effect of the operation using the columella instead of the broken incus was estimated. The shape, mounting position to the malleus，and the material of the columella were variously altered in the analysis. It was ascertained that hearing recovery of about 92% could be expected. Furthermore, harmonic vibration analysis of the middle ear’s ability to receive sound pressure was carried out in order to obtain the frequency response characteristics. It was determined that hearing recovery of about 98% could be expected. From the viewpoint of the static and dynamic analyses, it was proven that prediction of a hearing restoration effect was possible by our method, which made the displacement of the stapes basal plane to be a standard. These results are appropriate from a clinical viewpoint.


1. Introduction

The geometric model of the middle ear including the tympanic membrane, tympanic cavity, auditory ossicles, several ligaments, and tensor was constructed using SolidWorks®. The method for estimating the hearing restoration effect from the perpendicular displacement to stapes basal plane has been proposed in our research group. In order to verify its validity, the finite element


14 Takao HIGASHIMACHI and Ryuzo TORIYA

model in which the column article called columella is substituted for the incus of the healthy ear is constructed and the static analysis was performed in our previous research (Higashimachi, et al. 2013).

In this paper, a geometric model of the middle ear in which the ossicular chain is broken by the

middle ear cholesteatoma, is constructed. For this model, the finite element static analysis is carried out by a change of material, shape and the way of mounting the columella. From this, a prediction is made of the degree of hearing ability recovery and the validity of the prediction method is verified with the clinical data.

In actuality, the middle ear transmits the wave of the air vibration received from the tympanic

membrane to the inner ear through auditory ossicles. In the internal ear, the stapes vibration is transmitted to the labyrinthine fluid in the cochlea where electrical signals are generated. Finally, it is recognized in the brain as sound. Therefore, harmonic vibration analysis of the middle ear was done in order to examine the frequency response characteristics of the stapes. On the assumption of the case in which the part of the auditory ossicles was deficient, the rectangular part, called the columella, is installed between the malleus and the stapes. In this study, the possibility of clinical application of the prediction method of the hearing restoration effect is investigated from the viewpoint of the vibration analysis, too.

Figure 1: Ear structure.

2. MiddleEarStructureanditsFunction

2.1 MiddleEarStructure

Figure 1 shows the structure of the ear. The middle ear is composed of the tympanic

Pinna

Ear canal

Tympanic membrane

Auditory tube

Auditory ossicles

Cochlea

Semicircular


15

membrane, tympanic cavity, auditory ossicles and others which are shown in Figure 2. Auditory ossicles are behind the tympanic membrane in a small space(tympanic cavity) , and they are composed of malleus, incus and stapes. The stapes basal plane connects with the inner ear through an oval window.

Figure 2: Auditory ossicles (modified from: http://toppatu.com/sikumi_mimi_ototutaeru.html)

2.2 FunctionofTympanicCavity

The tympanic cavity is a space filled with air. The inner wall of the tympanic cavity is covered

with a mucous membrane. The air pressure of the tympanic cavity is controlled at the appropriate

value by ventilation in order to keep the important function that is “the sense of hearing”.

Furthermore, the tympanic cavity has a washing function that absorbs and removes bacterial waste

by the secretion and reabsorption of the mucus.

2.3 FunctionofAuditoryOssicles

Each part of the auditory ossicles is connected with the joints. They are suspended by

ligaments and muscles in the tympanic cavity. The vibration amount of the tympanic membrane is

amplified about 17 times by the area ratio of the stapes basal plane and the tympanic membrane. In

addition, the vibration amount of the tympanic membrane is amplified about 1.3 times by the lever

motion of the auditory ossicles.

2.4 AuditoryOssiclesMovement

Figure 3 shows that the auditory ossicles turn about the axis connecting the superior mallear

ligament and the posterior incudal ligament. By this rotary motion, the vibration of the tympanic


membrane is efficiently converted into Z-direction (perpendicular to the stapes basal plane)

movement of the stapes. The stapes vibration is transmitted to the labyrinthine fluid of the internal

ear and converted to electrical signals, which are then recognized as sound in the brain. In this

study, we consider that there is a certain relationship between hearing ability and the Z-direction

displacement of the stapes.

Figure 3: Deformation of auditory ossicles.

3. StaticAnalysisApproach

3.1 FiniteElementModeling

3.1.1 BrokenOssicularChainModel

Figure 4 shows the geometric model of the middle ear in which the ossicular chain is broken

by the middle ear cholesteatoma. The cholesteatoma adhered to the part bounded by the red circle

in Figure 4. Therefore, a connection between the incus and the stapes is lacking.

Figure 4: Broken ossicular chain model

Direction ofrotation

Superior mallear ligament

Axis of rotation

Posterior incudal ligament

Movement


17

3.1.2 TympanoplastyModelforBrokenOssicles

The geometric model for the finite element analysis is shown in Figure 5. This model is

composed of 10 parts. The CT scanning data of the human head were converted to DICOM data,

and subsequently, converted into STL data which were imported into SolidWorks.

The malleus normally does not work, because it is adhered to the neighboring bones by the

cholesteatoma. Therefore, the upper part of the malleus was detached from the anterior mallear

ligament and superior mallear ligament. In a clinical operation, the columella is installed between

the malleus and the stapes instead of the damaged incus. This operation method is called the

“Auditory ossicles formation Ⅲ-i type”. The schematic view (Morimitsu 1979) and geometric

model sample of the Ⅲ-i type formation are shown in Figure 6. The material data of the columella

are the same as silicon or human bone, and the joint part as cartilage.

Four kinds of models that change the mounting location of the columella were analyzed in

order to investigate the difference of the displacement of the stapes basal plane. These geometric

models are shown in Figure 7. On the right side of this figure, the columella is connected to the

tympanic membrane, too. Material data of each part are shown in Table 1. The numbers next to the

anatomical names in the table corresponds to the numbers in Figure 5.

Figure 5: Geometric model of middle ear.

A

①

④

②

⑤

⑥

⑩

⑧

③

⑦

⑨


(a) Schematic view (b) Geometric model sample

Figure 6: Ⅲ-i type formation model.

(a) Connection to umbilical part of malleus

(b) Connection to intermediate part of malleus

Figure 7: Geometric model for tympanoplasty.

Malleus

Columella

Stapes

Columella

Stapes

Tympanic


19

Table 1: Material data.

Anatomical name Young’s Modulus [MPa]

Poisson’s ratio

①Tympanic cavity wall 13,436

0.3

②Malleus ③Stapes ④Tympanic membrane 64 ⑤Anterior mallear ligament

21 ⑥Lateral mallear ligament ⑦Stapedial annular ligament 0.65 ⑧Stapedial muscle 0.52 ⑨Columella

Bone 13,436 Silicon 112,400 0.28

⑩Joint 6 0.3

3.1.3 BoundaryCondition

A total of 3 cross sections (one section is A of Figure 5) of the tympanic cavity were perfectly

fixed in constraint conditions. The sound pressure of 120 dB was converted into pressure using the

following equation as load conditions.

Lp =20log10(P/ P0) (1)

Where,

Lp =120dB: the limit of the sound pressure which the human auditory sense is able to hear

safely.

P0 =20×10-6 Pa: standard value (the lowest value of sound intensity which is audible for

humans).

As a result, the pressure of P=20Pa was given at the contact surface of the tympanic membrane

and malleus.

3.2 FiniteElementAnalysisResults

Figure 8 shows the displacement of the stapes in the Z-direction, which is perpendicular to the

basal plane for the sound pressure of 120 dB. In these results, the columella is connected to the

umbilical part of the malleus (Figure 7(a)) and its material composition is silicon. Table 2 shows


the average displacement of the stapes basal plane in each model.

Figure 8: Z-direction displacements of stapes (Umbilical connection model).

Table 2: Average displacement of stapes. Connection position

of columella Displacement [10-6mm] Silicon Bone

Umbilical part 2.70 2.67 Umbilical part & Tympanic membrane 2.22 2.22

Intermediate part 2.83 2.78 Intermediate part & Tympanic membrane 2.19 2.12

After deformation

Before deformation

Before deformation

After deformation


21

3.3 Considerations

3.3.1 MountingPositionofColumella

Typical methods of columella connection between malleus and stapes in Ⅲ-i type operation

are shown in Figure 9.

Figure 9: Typical connection methods of columella in Ⅲ-i operation.

In “overlay position” the columella is mounted over the short process of the malleus. In

“underlay position” the columella is mounted under the short process of the malleus. In “umbilical

position” the columella is mounted at the tip of the malleus. In “intermediate position” the

columella is mounted between the short process and the malleus tip.

It is easy to mount on the overlay and underlay position of the short process. However, the

head of the malleus is connected to the incus which is easier to damage by the middle ear

cholesteatoma or chronic otitis media among others. The short process is also easy to damage.

Therefore, it is difficult to mount the columella around the short process.

Intermediate

Short process

Umbilical

Head

Intermediate

Overlay Underlay

Umbilical


On the other hand, it is very effective to mount the columella on the umbilical position of the

malleus. However, the tympanic membrane can easily be caused inflammation by this mounting

process. Furthermore, the tip of the malleus is so small that it is very hard to fit the columella to the

umbilical position. In general, a method to mount the columella between the tip and the short

process of the malleus is adopted.

3.3.2 HearingRestorationEffects

Figure 8 shows that the stapes deform in a perpendicular direction to the basal plane, that is, in

the Z-direction. In this study, we consider that there is a certain relationship between hearing ability

and Z-direction displacement of the stapes basal plane. For a healthy subject, the average

displacement of the stapes for a sound pressure of 120 dB is 3.09 nm, which becomes a standard

value in our study (Higashimachi, et al. 2013).

Hearing restoration effects are discussed below using the results of Table 2. When the

columella made of silicon is attached to the intermediate portion of the incus, a maximum

displacement of 2.83 nm occurs. This value is about 91.6% of the healthy subject. The

tempanoplasty model, in which the middle ear is damaged by the middle ear cholesteatoma, can be

sufficiently recovered by the operation of “Auditory ossicles formation Ⅲ-i type”. This result is

appropriate from a clinical viewpoint.

When the columella made of silicon is attached to the umbilical part of the incus, the average

displacement is 2.70 nm. When the columella made of human bone is attached to the intermediate

portion or the umbilical part of the incus is 2.78 nm or 2.67 nm, respectively. In these cases, the

effect decreases slightly, but a recovery of hearing ability of over 85% can be expected.

Furthermore, when the columella made of silicon is attached to both of the umbilical and the

tympanic membrane, the average displacement is 2.19 nm (intermediate part of the incus) or 2.22

nm (umbilical part of the incus). In these cases, the effect is considerably lowered. The vibration of

the tympanic membrane can not sufficiently transmit to the stapes using the tympanic membrane

with its low rigidity. This corresponds with the vibration of the tympanic membrane of Ⅲ-i

tympanoplasty model (the columella is attached to the malleus) is easier to reach the stapes than Ⅲ

-c tympanoplasty model (the columella is attached to the tympanic membrane) (Higashimachi, et

al. 2013). This schematic view (Morimitsu 1979) and geometric model sample of Ⅲ-c type


23

formation are shown in Figure 10.

(a) Schematic view (b) Geometric model sample

Figure 10: Ⅲ-c type formation model.

(a) Healthy type model (b) Ⅲ-i type tympanoplasty model

Figure 11: Geometric model of middle ear.

4. DynamicAnalysisApproach

4.1 GeometricModeling

The geometric model of a healthy middle ear for the finite element analysis is shown in Figure

11(a). This model is composed of 14 parts. The CT scanning data of the human head were

converted to DICOM data, and subsequently converted into STL data, which were imported into

SolidWorks®.

Malleus

Columella

Stapes

Columella

Stapes

Tympanic membrane


On the other hand, Figure 11(b) shows the operation model called “Auditory ossicles

formation Ⅲ-i type”. In this model, the rectangular part called the columella is installed between

the malleus and the stapes instead of the damaged incus.

Table 3: Material data.

Anatomical name Young's Modulus

[MPa]

Density [kg/m3]

Poisson's ratio

①Tympanic membrane 33.4 1,200

0.3

②Malleus 13,436 4,350 ③Incus

④Stapes ⑤Lateral mallear ligament

21

2,500

⑥Superior mallear ligament ⑦Anterior mallear ligament ⑧Posterior incudal ligament

0.65 ⑨Superior incudal ligament ⑩Stapedial annular ligament ⑪Incudostapedial joint

6 1,200 ⑫Incudomallear joint ⑬Stapedial muscle 0.52 2,500 ⑭Base plate 1×1010 ‐

⑮Columella(Silicon) 112,400 2,330 0.28 ⑯Joint 6 1,200 0.3

4.2 FiniteElementModeling

4.2.1 MaterialSetting

Material data of each part are shown in Table 3. The anatomical name numbers in the table

corresponds to the numbers in Figure 11. These data are determined by referring to the research of

Higashimachi et al. (2013) and Koike et al. (2002). The base plate is a virtual part for supporting

the spring. Therefore, its Young’s modulus can be assumed to be a rigid body. The columella is

composed of the main body and the joint. Silicon was used for the main body, and cartilage for the

joint.

4.2.2 BoundaryCondition

The attachment portion between the tympanic cavity and the tympanic membrane, muscle, and


25

ligament was perfectly fixed in constant conditions. Furthermore, the base plate was perfectly

fixed. The sound pressure of 90 dB was converted into pressure using the equation (1) as load

conditions. In equation (1), Lp =90 dB is the relative noisy sound pressure and P0 =20×10-6 Pa is

standard value. As a result, a pressure of P=0.632 Pa was obtained. However, in this analysis,

P=15.2 Pa was given at the contact surface of the tympanic membrane and malleus. The ratio

15.2/0.632 equals the ratio of the total area of the tympanic membrane to the contact surface area of

the tympanic membrane and malleus.

The spring of 40N/m spring constant was installed between the stapes and the base plate

referring to the research of Gan et al. (2004). Rayleigh damping was assumed and a damping factor

α = 0 s-1 and β = 7.5×10-5 s.

Figure 12: Frequency response graph of healthy subject

4.3 FiniteElementAnalysisResults

In this research, the harmonic vibration analysis was done as dynamic analysis using the finite

element method. Figure 12 shows the harmonic vibration analysis results of a healthy subject.

Figure 13 shows the comparison of some frequency response graphs. The solid line in Figure 13

shows the frequency response graph of our harmonic vibration analysis. In Figure 12 and Figure

13, the longitudinal axis shows the average displacement of the stapes bottom in the Z-direction

which is perpendicular to the basal plane, and the lateral axis represents the frequency.


The thin solid line and the dotted line in Figure 13 are the measurement results from the

research of Gan et al. (2001). The dashed and single-dotted line is the finite element analysis result

of Sun et al. (2002). The dashed and double-dotted line is the finite element analysis result of Koike

et al. (2002). Figure 14 shows the frequency response graph of the stapes bottom of our harmonic

vibration analysis in auditory ossicles formation Ⅲ-i type.

Figure 13: Comparison of frequency response graphs.

Figure 14: Frequency response graphs of tympanoplasty Ⅲ-i model.

4.4 Considerations

In cases of a healthy subject, it is said that there is a resonance region of the middle ear at 1～


27

2 kHz of frequency. Average displacement of the stapes basal plane shows the peak of the

resonance to be near 1.3 kHz of frequency in the analytical results of a healthy subject. This

displacement decreases gradually with an increase in frequency over 2 kHz. The measurement

result of the research by Gan et al. (2001) shows the peak of resonance to be near 0.7 kHz in Figure

13. This data is a mean value of 17 examinees and the dispersion between individuals is also large.

Therefore, it was possible to reproduce the resonance phenomena to some extent by our finite

element model. However, the average displacement of the stapes bottom is relatively smaller than

the measurement result of Gan et al. (2001) or the analysis result of Sun et al. (2002) and Koike et

al. (2002) under 1 kHz in Figure 13. Reasons for this could be as follows:

(a)Some of the material properties such as the ligaments are estimated values.

(b)The tympanic membrane has been divided into regions, and the material properties differ in

each regions. However, in our study, the membrane is treated as one part.

Figure 14 shows that the position of the peak of resonance does not change from about 1.5

kHz, even if the attaching position of the columella is changed in the Ⅲ-i type tympanoplasty

model. However, the average displacement of the stapes has the largest value at the position in

which the columella is attached at the umbilical region, that is, the tip of the malleus. The value

of the average displacement is 5.0 nm at about 1.5 kHz of frequency. Figure 15 shows the malleus

rotate around the superior malleus ligament. Therefore, the closer the attach point of the colemella

gets to the umbilical region -that is, the far from the superior mallear ligament-, the larger the stapes

displacement becomes.

Figure 12 shows the maximum displacement of the healthy type is 5.1 nm. On the other hand,

the maximum displacement of 5.0 nm of the Ⅲ-i type model in Figure 14 is about 98 % of the

healthy type. This means that hearing ability recovers to the normal level. We have proposed that

the hearing restoration effect can be estimated by the displacement of stapes basal plane. Its

validity has been verified by static finite element analysis. The same statements are true for the

harmonic vibration analysis. Therefore, it becomes possible to estimate the restoration ratio by

comparing the stapes displacement of the tympanoplasty model with the healthy type in all range of

the frequency.


Figure 15: Vibration mode of Ⅲ-i model (Frequency f=1.7kHz).

4.5 Conclusions

We have already proposed the method for estimating the hearing restoration rate by comparing

the displacement of the stapes basal plane (Higashimachi, et al. 2013). In this study, the validity of

our method was verified by finite element static and dynamic analysis.

On the actual patient whose ossicular chain was broken by the middle ear cholesteatoma, the

tympanoplasty operation was simulated and the hearing restoration effect was estimated according

to our method. When the columella made of silicon is attached to the intermediate portion of the

incus, a maximum displacement of 2.83 nm occurs. This value is about 91.6% of a healthy subject.

This result is appropriate from a clinical viewpoint.

By harmonic vibration analysis, the frequency response characteristics of a healthy subject

were clarified. The displacement of the stapes basal plane shows the peak of the resonance to be

near 1.3 kHz of frequency. This displacement decreases gradually with an increase in frequency

over 2 kHz.

On the assumption of the case in which the incus of the healthy subject was damaged,

harmonic vibration analysis was applied to this broken model. As a result of analysis, hearing

ability recovers to the normal level -that is, about 98% of a healthy subject- by the Ⅲ-i type

tympanoplasty model.

We have proposed that the hearing restoration effect can be estimated by comparison of the

Before deformation

Direction of rotation

mallear ligament Superior

After deformation


29

displacement of stapes basal plane. The validity of our proposal was confirmed by the static and

dynamic analysis approach. This kind of approach makes it possible to propose a new medical

treatment for the recovery of conductive hearing loss.

5. References

Gan, R. Z., B. Feng and Q. Sun. ( 2004). Three-dimensional finite element modeling of human ear for sound transmission. Annals of biomedical engineering, 32, 847-859.

Gan, R.Z., R.K. Dyer, M.W. Wood, and K.J. Dormer. (2001). Mass loading on ossicles and middle ear function. Ann Otol Rhinol Laryngol 110, 478-485.

Higashimachi, T., Y. Shiratake, T. Maeda, and R. Toriya. (2013). Three-dimensional finite element analysis of the human middle ear and an application for clinics for tympanoplasty. Surface Effects and Contact Mechanics, XI, WIT Transactions on Engineering Sciences, Vol. 78, 61-72.

Koike, T. and H. Wada. (2002). Modeling of the human middle ear using the finite-element method. Journal of Acoustical Society of America, 111(3), l306-1317.

Sun, Q., R.Z. Gan, K.-H. Chang, and K.J. Dormer. (2002). Computer-integrated finite element modeling of human middle ear. Biomeshan Model Mechanobiol, 1, 109-122.

Morimitsu, T. (1979). Illustrated Ear Surgery, Medical Illust Co., Tokyo, Japan.

Dr. Takao Higshimachi is a Professor of Mechanical Engineering at Sojo University, Kumamoto, Japan. His research interests include investigating the application of CAD/CAE to biomechanics in extending the longevity of the teeth and ear. At present, he is working on the development of a new prediction system of the hearing restoration effect using the finite element method.

Dr. Ryuzo Toriya is a Medical Doctor at Toriya ENT clinic, in Kumamoto, Japan. He is an ear and nose surgeon with a particular interest in developing ear surgery using the finite element method.

Note: The original of this article has been submitted to The 3rd International Conference on Design Engineering and Science (ICDES 2014), held at Pilsen, Czech Republic. The Paper Award Committee of ICDES 2014 has reviewed and selected this paper for journal publication.



Call-for-Papers:





*Corresponding author (Krittiya Lertpocasombut). Tel: +66-2-5643005 x3109. E-mail: [email protected]. 2015. American Transactions on Engineering & Applied Sciences. Volume 4 No.1 ISSN 2229-1652 eISSN 2229-1660 Online Available at http://TUENGR.COM/ATEAS/V04/031.pdf.

31



Management of Garbage Problems with Participation of the 21 Dormitory Buildings' Community

Krittiya Lertpocasombut a*

, Boonsap Witchayangkoon a

and Sayan Sirimontree a

a Department of Civil Engineering, Faculty of Engineering, Thammasat University, THAILAND.


A B S T R A C TArticle history: Received 30 September 2014 Received in revised form December 24, 2014 Accepted 27, 2014 Available online January 05, 2015 Keywords: Community participation; Questionnaire survey; Dormitory resident; Solid waste; Study area.

When the community grew as the population increased, it was resulted in waste increased. The problems of solid waste affect the environment and people's health. The study area of a community of 21 dormitory buildings near Thammasat University Rangsit Campus, the residential area has found rubbish overload from the tank waste, provoked smell and being a source of disease carriers such as rats and roaches. For this reason, the residents should bear in dealing with solid waste that occurs without dependence on government unilaterally in waste disposal. In addition, the rapid growth in the number of dormitories surrounding Thammasat University Rangsit Campus has ignited in an attempt to fix the problem, garbage overload. Questionnaire survey of the residents showed up with a random sampling method. The residents who live in buildings affected by high, medium, and low impacts of garbage overload tanks are chose for surveying questionnaire. In addition, the store owners and the manager of the buildings are surveyed to find out their opinions and recommendations on garbage campaign management. The results showed that most of the residents had been not satisfied with the management of garbage disposal. They would like to have an exchange of solid waste recycling as a reward rather than in cash and are willing to cooperate with the campaigns when they could earn some kind benefits.


1. Introduction

A local community’s garbage disposal has been expanded to a current situation due to a

population increase. This provoked problems of garbage spilled out which impacted on the

environment and affected also health of the public. Therefore, people who live in the community


32 Krittiya Lertpocasombut, Boonsap Witchayangkoon, and Sayan Sirimontree

should pay attention on the contribution of solid waste management rather than dispose garbage

away from their backyards. Therefore, it should not be unilaterally depended on the local

government management.

A case study was selected an area nearby Thammasat University Rangsit Campus with a

number of 21 resident buildings, so-called “the 21 Dormitory Buildings’ Community”. There are

three types of the dormitory buildings: type A, B, and C. Each building of type A and type B

consists of 8 floors with 19 dormitory rooms per floor. Type C has 8 floors with 33 dormitory

rooms per floor. A total number of the residents is estimated to be 6,000 persons in year 2015.

Garbage generated from this community would reach 6.6 Tons a day, estimated with 1.1 kg/capita

day (Zurbrugg, C., 2002). Meanwhile, the local government has not formulated measures for

garbage disposal to serve further fast development of the community.

Following this considerations, the research study aims to investigate garbage generated from

the community of 21 dormitory buildings due to its future problems faced on garbage management.

For this reason, and also it comes to our social responsible considerations within the environmental

concerns surrounding of Thammasat University Rangsit Campus, Pathumthani province.

2. Methodology

2.1 CurrentGarbageSituationsandProblems

Garbage overloading tanks have appeared where the garbage storage allocated in the

community. It provoked smell and being a source of disease carriers such as rats and roaches

including wastewater occurred in the surrounding area. By observing, the garbage tanks are not

sufficient due to mal design, since, dormitory rooms are available. Some of the buildings (C2 and

C8) are recently construction finished. There are areas provided for shop and food court in the

plaza, and a building for parking as well. Plan view of the community is shown in Figure 1.

2.2 SatisfactionandParticipationofGarbageCampaignManagement

This work uses questionnaire as a tool to survey individual dormitories regarding their

satisfactions of the buildings’ community management on garbage disposal and their willingness to

participate to which type of campaign management of garbage recycling.

2.3 QuestionnaireSurvey

A questionnaire consists of a series of questions (open-ends or closed-ended questions) which


33

divided into 3 parts. Those are part I: Respondent general information; part II: Management of

garbage disposal including knowledge, behavior, and profitability concerned on garbage recycling;

part III: Recommendations on preferable garbage campaign.

Figure 1 Plan view of the 21 dormitory buildings’ community

A pre-test surveying was handed out to 10 respondents, one respondent per dormitory. In the

questionnaire, individual recycling behaviors were divided into routine, seldom and never

practices, as defined in the following:

Routine: do collect as a question posed on daily basis. Seldom: do collect as a question posed in once on several days. Never: do not collect as a question posed.

2.4 InclusionCriteriaandSampleSize

In this study, random method is used for questionnaire survey of 98 respondents as a sampling

number based on Yamane’s formula (1967). The residents who live in buildings affected by high,

medium, and low impacts of garbage overload tanks are chose for surveying questionnaire. In

addition, the store owners and the manager of the buildings in study area of “the 21 Dormitory

Buildings’ Community” are surveyed to find out their opinions and recommendations on garbage

campaign management.

3. ResultsandDiscussion

Respondents classified by gender, ages, and education levels in order to compare among the


store owners and the manager of the buildings are illustrated in Table 1. Most of the respondents

are ages ranged from 20 to 30. There are 98 respondents of 98 sampled residents, giving a 100 %

response rate. Of these 98 respondents, 55 women and 43 men responded to a face-to-face

questionnaire. Survey data collected showed that 84 % were educated in undergraduate level and

16 % were in graduate level.

Table 1 Respondents classified by gender, age, and education compared to store owners & manager

Number of Classification Residents Store Owners Manager

Gender Female Male Total (percent)

55 43

98 (100)

5 0 5

0 1 1

Age < 20 20 to 30 > 30 Total (percent)

0 98 0

98 (100)

0 0 5 5

0 0 1 1

Education College/Training Undergraduate Graduate Others Total (percent)

0 82 16 0

98 (100)

5 0 0 0 5

0 0 1 0 1

Income per month (Baht) Low (<10,000) Lower middle (10,000 to 15,000) Middle (15,001 to 20,000) Upper middle (>20,000) High (>30,000) Total (percent)

6 28 55 9 0

98 (100)

0 0 0 3 2 5

0 0 0 0 1 1

On the basis of monthly disposable income per capita from the surveyed residents, most of

residents are lower-middle income (10,000 to 15,000 Baht) and middle income (15,001 to 20,000

Baht). Note 30 Baht is about 1 U.S. Dollar.

3.1 RecognitiontowardtheParticipationofGarbageCampaignManagement

Majority of the residents have not approved of classified garbage collection (Table 2). On an

average of 56 % of the residents do not understand the significant of classified garbage collection

and are not willing to classify their daily garbage before disposing it. But only 20 % on the

average are willing to practice classified garbage collection in their dormitory rooms. This figure

of recognition of the residents is much higher than those of the store owners, and the manager in the


35

buildings’ community, respectively.

Table 2 Recognizing of residents, store owners, and manager towards the classified garbage collection.

Participation Rate, % Residents Store Owners Manager Routine collect 20 0 0 Seldom collect 24 60 0 Never collect 56 40 100

Total (percent) 98 (100) 5 1 For economic incentive, residents in low income are willing to collect recyclable materials.

This result is corresponding to the recycling behavior in a big city in China that individuals in lower income households were more active in recycling (Li, S., 2003). In Korea household survey data from Hong, S. (1999) indicates that a rise in waste collection fee induces households to recycle more wastes. In addition, more frequent recyclable pickup services accompanied with increasing in waste collection fee is effective to reduce total amount of waste generated.

Our study survey shows that all residents (100 % of respondents) would like to have an exchange of solid waste recycling as a reward rather than in cash. And almost all of the residents (91 % of respondents) prefer the building owner to conduct various activities in garbage campaign with regard to their flexibility to participate. Thus, the residents there can join a few policymaking oriented suggestions for transferring the strategy of garbage management from passive control to active source control and promoting the classified garbage collection.

3.2 Respondents’OpinionsontheExistingGarbageManagement

The majority of respondents are unhappy with the garbage tanks’ area of the community where is a place of disease carriers like rodents, cockroaches, flies and mosquitoes. Other problem issue of the community related to solid waste is the overflow of garbage from the tank. Whereas, respondents satisfied with the environment is minimal. Because those respondents who are affected by unfavorable environment are not seen as a problem bothers.

Some residents’ opinions based on profitability are wishing to see the development of community in a better way. They desire to have profitability from solid waste recycled and realize that each one is a part of the community. Some of them urge to follow such garbage campaigns as soon as possible and wish to join as well. In a contrast opinion, one could not trust on the building owner to conduct any garbage campaigns because he believes that the building owner does not pay attention on the garbage management.


The manager of the buildings responds that the garbage collection services are provided 3

times a week by local government. This is usually a problem of any operating offsite the

buildings’ vicinity. The problems provoked herein are the less budget concerned to garbage

disposal, the lack of knowledge about solid waste management, and the building design has not

taken into account the environmental criterion. The waste problem in the buildings’ community

was similar to the case of Rayong Municipality (Kritjaroen, T., 2011) due to rapid increases in

population; lack of proper disposal units; limited budget and landfill areas.

The store owners are willing to cooperate in the garbage campaign to reduce solid waste and

help the environment clean because they believe that it is better to trade and service with a clean

environment. When stores keep sanitary, people in the community has confidence and comes to

shop more. On the other point of view, the store owners are worried that residents in the

community would not cooperate in reducing the use of foam food boxes because of the

convenience of carrying foam food boxes to their rooms.

4. Conclusion

This study has met its aim of understanding some of the opinions from the residents, the store

owners and the manager of the 21 dormitory buildings’ community. The residents are not

satisfied with existing garbage management. And they declare cooperation with any campaigns

on the conditions that the administrative of the residence has organized to return some income or

benefits in various ways from the solid waste campaign. The residents believe that their own

cooperation by working together can make a clean environment but still worry the building owner

does not really make the campaign happen.

The study reveals that the manager of the buildings does not understand the problem of

garbage management. The manager also does not pay attention to garbage management that

would contribute to a quality of good livable community, either in short-term or long-term. On

the other hand, the store owners are willing to participate and support such garbage campaign

management in order to get a livable community.

5. Acknowledgements

The authors would like to appreciate Miss Natcha Permpool and Miss Orapin Jittanupat from Engineering and Business Management Program (EBM), Faculty of Engineering, Thammasat University, for their time in helps managing this questionnaire survey. Many thanks also go to the


37

residents of “the 21 Dormitory Buildings’ Community” who participated in this study.

6. References

Hong, S. (1999). The effects of unit pricing system upon household solid waste management: The Korean experience. Journal of Environmental Management, 57(1), 1-10.

Kritjaroen, T. (2011). Understanding urban governance in the context of public-private partnerships: A case study of solid-waste management in Rayong Municipality, Thailand. Federal Governance, 8(3).

Li, S. (2003). Recycling behavior under China’s social and economic transition: The case of metropolitan Wuhan. Environment and Behavior, 35(6), 784-801.

Zurbrugg, C. (2002). Urban solid waste management in low-income countries of Asia how to cope with the garbage crisis. Presented for: Scientific Committee on Problems of the Environment (SCOPE) Urban Solid Waste Management Review Session, Durban, South Africa, November, 2002, 1-13.

Dr.Krittiya Lertpocasombut is an Associate Professor in the Department of Civil Engineering, Faculty of Engineering, Thammasat University, Thailand. She received a B.Sc. in Chemistry from Chulalongkorn University, Thailand, an M.Sc. in Environmental Engineering, Asian Institute of Technology (A.I.T.), D.E.A. Diplome d’Etudes Approfondies in Water Purification and Treatment Engineering from INSA de Toulouse, France, and a PhD in Water Purification and Treatment Engineering, Institut National des Sciences Appliquees (INSA), Toulouse, France. Dr. Lertpocasombut is interested in water and wastewater treatment; wastewater recycled by membrane technology; water supply sludge treatment and its reuse/recycle.

Dr. Boonsap Witchayangkoon is an Associate Professor of Department of Civil Engineering at Thammasat University. He received his B.Eng. from King Mongkut’s University of Technology Thonburi with Honors in 1991. He continued his PhD study at University of Maine, USA, where he obtained his PhD in Spatial Information Science & Engineering. Dr. Witchayangkoon current interests involve applications of emerging technologies to engineering.

Dr. Sayan Sirimontree earned his bachelor degree from Khonkaen University Thailand, master degree in Structural Engineering from Chulalongkorn University Thailand and PhD in Structural Engineering from Khonkaen University Thailand. He is an Associate Professor at Thammasat University Thailand. He is interested in durability of concrete, repair and strengthening of reinforced and prestressed concrete structures.

Note: The original of this article has been submitted to 2nd International Workshop on Livable City 2014 (IWLC2014), a Joint Conference with International Conference on Engineering, Innovation, and Technology (EIT), held at Tabung Haji Hotel, Alor Star, Malaysia, during December 9-11, 2014. According to the IWLC2014 Conference Committee, this paper was given Best Presentation Award.



Call-for-Papers:





*Corresponding author (Sayan Sirimontree). Tel: +66-2-5643006 x3112. E-mail: [email protected]. 2015. American Transactions on Engineering & Applied Sciences. Volume 4 No.1 ISSN 2229-1652 eISSN 2229-1660 Online Available at http://TUENGR.COM/ATEAS/V04/039.pdf.

39



Strengthening of Reinforced Concrete Column via Ferrocement Jacketing

Sayan Sirimontree a*

, Boonsap Witchayangkoon a

and Krittiya Lertpocasombut a



A B S T R A C TArticle history: Received 26 September 2014 Received in revised form December 22, 2014 Accepted December 27, 2014 Available online January 05, 2015 Keywords: Ferro cement; strength; Ductility; ACI; wire mesh; cement mortar; steel rebar.

This work focuses on behaviors of reinforced concrete (RC) column encased by longitudinal steel and ferro cement under static axially loading. RC column specimens are encased by vertical steel reinforcements, wrapped by varying amount of wire mesh and then covered with cement mortar. The results show significantly improvement of strength and ductility of strengthened column over reference column without strengthening. Ductility is also significantly improved by the increase of the volume of wire mesh. ACI equation for prediction of strength of short axially loaded RC column can be applied to predict strength of both reference and strengthened column.


1. Introduction

Structural building components in aged RC buildings need to be investigated and

maintenances due to the deteriorations of concrete with time. Deterioration of concrete can be

accelerated by aggressive environment and low quality of concrete as can be seen in Figure 1a.

Low quality of concrete leads to high porosity which moisture can penetrate to concrete pore result

in corrosion of steel in concrete. Spalling of concrete cover caused by rust from corrosion process

has higher volume than original steel for about 4 times. Corrosion rates of steel increase rapidly

with time if repair action is not performed. Reduction of concrete and steel area cause load


40 Sayan Sirimontree, Boonsap Witchayangkoon, and Krittiya Lertpocasombut

carrying capacity of structural elements to decrease. Not only structural deterioration but also the

need of changing functional use of building may cause increasing service load of building

components.

(a) Corrosion of steel in concrete (b) Repair by nonshrink mortar and

strengthening by steel jacketing

Figure1: Degradation of RC column and column strengthening.

RC column can be classified as the most important component of the building super structure

since load from slabs and beams are both transferred to columns. Total collapse of RC building

may occur due to change of service load and lack of column strength caused by deterioration.

Repair and strengthening to increase load carrying capacity of column can be performed by distill

degraded concrete, patch by nonshrink mortar and then strengthened by steel jacketing (Figure 1b)

or encased by additional RC. Ferro cement jacketing is one of the alternate method of repair and

strengthening of column which is low cost and easy to apply to existing column, as do-it-yourself

(DIY). As describe above, behaviors of columns strengthened by additional steel reinforcement

and encased by Ferro cement under static loading are studied in this work.

Figure 2: Example of steel wire mesh (Paul, 2013).

2. Ferrocement

ACI Committee 549, 1997 state that “Ferrocement is a type of thin wall RC, commonly

constructed of hydraulic cement mortar, reinforced with closely spaced layers of continuous and

relatively small size wire mesh (see Figure 2). The mesh may be made of metallic or other


41

suitable materials.”

Advantages of ferrocement are high ductility, reduce number of cracks and crack width, high

deformation capacity, improve impact resistance and toughness, good fire resistance, low

permeability, low cost of maintenance and high strength to weight ratio.

3. ReviewofLiterature

Ferrocement has great uses and applications (Naaman, 2000). There are many research on

strengthening structural members by ferrocement.

Flexural analysis and behavior of ferrocement beam was studied by Balaguru et al. (1977).

The study was able to predict ultimate flexural capacity and deflections and crack widths

ferrocement beams under loading. The study also observed the load-deflection curves, crack

distribution, and crack widths for ferrocement beams up to ultimate. From test measurement of

stress-strain curves for mortar in compression (including the descending portion) and steel meshes

in tension, an analytical model was developed to generate the nonlinear moment-curvature and

load-deflection curves of ferrocement beams.

Moment capacity and cracking behavior of ferrocement beam in flexure was studied by Logan

and Shaw (1973). Results of tests on ferro-cement beams was presented with data on initial

cracking, crack widths and ultimate strength. Ferrocement beams were compared with

conventionally reinforced concrete beam.

Strengthening of RC beams with ferrocement laminates was studied by (Paramasivam et al.,

1998). The test can be observed on the effects of the level of damage of original beams prior to

repair, and repeated loading on the performance of the strengthened beams. The study found that

ferrocement is a practical method to strengthening and rehabilitation of reinforced concrete

structures.

Flexural behavior of reinforced concrete beams with ferrocement thin plates reinforced with

steel wire mesh was experiment by Shang et al. (2003). The test comprised 13 RC beams

strengthened by steel wire mesh and 2 specimens without strengthening for comparison. The

strengthening results of reinforced with three side (U shaped ferrocement put onto the tension face


and two profile faces) and one side ferrocement had been analyzed. The study observed the

performance of the tested beams, the modes of failure. Also, the test measured mid span

deflection, crack width and strains of steel and wire mesh. The ferrocement can obviously increase

the load bearing capacity and crack resisting capacity, and improve the bending stiffness of beam.

Many researches and discussions are about testing beams with ferrocement, but there are little

researches on testing columns with ferrocement. Thus, this work will show behaviors of RC

columns strengthened by ferrocement.

Figure 3: Details of test specimens.

4. ExperimentalStudies

This experiment has six tested columns, one being referenced specimen (CR) and five being

tested columns encased by ferrocement (CF-7, CF-9, CF-11, CF-13, and CF-15). Referenced

specimen CR is a 150x150x1500mm-size column reinforced by 4DB12mm longitudinal steels and

RB6mm with stirrup spacing 150mm. All five columns strengthening with ferrocement ((CF-7,

CF-9, CF-11, CF-13, and CF-15) have the same core column and reinforcement details as the

referenced specimen CR. These columns are strengthened by additional longitudinal 4DB12mm

CR

CF-7,

CF-9, CF-11,

CF13, CF15


43

steels, and wrapped by square welded wire mesh, and put with cement mortar in order have

cross-sectional dimensions 300x300mm. The dimensions of all RC column specimen and

reinforcement details are shown in Figure 3. The numbers 7, 9, 11, 13 and 15 refer to number of

wrapping rounds of wire mesh around the existing column. Details of all test specimens are given

Table 1.

Ordinary Portland cement is used both in mixing of mortar and concrete. Water to cement ratio

(w/c) of mortar cement used in this work is 0.5. The average compressive strength of mortar from

test is 21.7MPa. Average concrete strength and yield strength of steel from test are 11.5Mpa and

327.1MPa respectively.

Test set up of column specimens can be readily seen by the diagram shown in Figure 5. Load

is applied to column statically or gradually increased by hydraulic jack. Load, strain and

deformation of column can be captured by data logger and transfer to computer.

Table1: Details of reference and strengthened specimens.

Figure 4: Wrapping of wire mesh before applying mortar cement.


Figure 5: Test set up of column specimen.

Figure 6: Relationship of load and displacement of test specimen.


Relationships of compressive load and displacement (contraction) of all test columns are given

in Figure 6. It can be observed that significant improvement of static strength and ductility of all

strengthened specimens over reference specimen CR. This is due to the additional area of column

section both ferrocement and reinforcing steels. The deformation capacity prior to failure which

indicates the ductility of column depends on the number of wrapping rounds of wire mesh. The

result shows optimum number of wrapping round of wire mesh is 13 (specimen CF-13).

Increasing of ductility is caused by the efficient confinement of wire mesh and mortar cement

composite.


45

It can be said that ferrocement is equivalent to RC but its advantage is higher ductility due to

the confinement of wire mesh composite with mortar cement. The prediction of nominal

compressive load of column should be used the modified ACI equation as shown by Equation (1).

' '0.85 0.85n c c s y cf cf sf yP f A A f f A A f (1)

where nP = Nominal loading capacity of column

'cf =Concrete compressive strength

'cff = Compressive strength of cement mortar

yf = Yield strength of steel

cA = Gross area of concrete

cfA = Area of cement mortar

sfA = Area of additional steel

Nominal load carrying capacities of all test columns comparing to load prediction Equation (1)

are exhibited in Figure 7. It can be seen that equation (1) can be applied to predict maximum

nominal load carrying capacity both RC column and RC column strengthened by additional

reinforcement and ferrocement.

Typical failure mode of reference column and strengthened column can be observed in Figure

8. It can be noticed that concrete core is prevented by ferrocement lead to increase of column

ductility (see Figure 6).

Figure 7: Nominal load (strength) of test specimens compared to load predictions Equation (1)

(dotted line).

0.0

500.0

1000.0

1500.0

2000.0

2500.0

CR CF-7 CF-9 CF-11 CF-13 CF-15

(,

)nP

kN

368.4 kN (Reference column)

1760.7 kN (Strengthened column)

Predicted load by equation 1


Figure 8: Failure mode of test columns.

6. Conclusions

Behaviors under concentrically static loading of RC column strengthened by additional

reinforcement and ferrocement are studied. The following conclusions can be made.

1) Strength of RC column can be significantly improved by additional steel and ferrocement.

2) Modified ACI equation can be used to predict static strength of both RC column and RC column strengthened by additional steel and ferrocement.

3) Concrete core can be prevented by ferrocement leading to high ductility of strengthened column.

7. Acknowledgements

Authors are grateful to Mr.Kan Kaewkaemket, Mr.Rachpong Rungrueangyingyod and

Mr.Anan Manokiang for their helps in setting up the experiment. This research is partially funded

by Faculty of Engineering, Thammasat University.

8. References

Paul, Anand. 2013. What is ferro cement. http://civildigital.com/what-is-ferro-cement. Accessed October 2014.

Naaman, A. E. (2000). Ferrocement and laminated cementitious composites (Vol. 3000, No. 1). Techno press.

Balaguru, P. N., Shah, S. P., & Naaman, A. E. (1977). Analysis and behavior of ferrocement in flexure. Journal of the Structural Division, 103(10), 1937-1951.

Paramasivam, P., Lim, C. T. E., & Ong, K. C. G. (1998). Strengthening of RC beams with


47

ferrocement laminates. Cement and Concrete Composites, 20(1), 53-65.

Logan, D., & Shaw, S. P. (1973, December). Moment capacity and cracking behavior of ferrocement in flexure. In ACI Journal Proceedings (Vol. 70, No. 12). ACI.

ACI Committee 549, State-of-the-Art Report on Ferrocement, ACI 549R-97, ACI Committee 549 Report, American Concrete Institute, Farmington Hills, Michigan, 1997.

Dr. Sayan Sirimontree earned his bachelor degree from Khonkaen University Thailand, master degree in Structural Engineering from Chulalongkorn University Thailand and PhD in Structural Engineering from Khonkaen University Thailand. He is an Associate Professor at Thammasat University Thailand. He is interested in durability of concrete, repair and strengthening of reinforced and prestressed concrete structures.

Dr. Boonsap Witchayangkoon is an Associate Professor of Department of Civil Engineering at Thammasat University. He received his B.Eng. from King Mongkut’s University of Technology Thonburi with Honors in 1991. He earned his PhD from University of Maine, USA in Spatial Information Science & Engineering. Dr. Witchayangkoon current interests involve applications of emerging technologies to engineering.

Dr. Krittiya Lertpocasombut is an Associate Professor in the Department of Civil Engineering, Faculty of Engineering, Thammasat University, Thailand. She received a B.Sc. from Chulalongkorn University, Thailand, an M.Sc. from Asian Institute of Technology, D.E.A. Diplome d’Etudes Approfondies in Water Purification and Treatment Engineering from INSA de Toulouse, France, and a PhD in Water Purification and Treatment Engineering, Institut National des Sciences Appliquees (INSA), Toulouse, France. Dr. Lertpocasombut is interested in water and wastewater treatment; wastewater recycled by membrane technology; water supply sludge treatment and its reuse/recycle.

Note: The original of this article has been submitted to 2nd International Workshop on Livable City 2014 (IWLC2014), a Joint Conference with International Conference on Engineering, Innovation, and Technology (EIT), held at Tabung Haji Hotel, Alor Star, Malaysia, during December 9-11, 2014. According to the IWLC2014 Conference Committee, this paper was given Technology Best Paper Award.



Call-for-Papers:





*Corresponding authors (Boonsap Witchayangkoon) Tel +66-2-5643005 Ext 3101 [email protected] 2015. American Transactions on Engineering & Applied Sciences. Volume 4 No.1 ISSN 2229-1652 eISSN 2229-1660 Online Available at http://TUENGR.COM/ATEAS/V04/049.pdf.

49



Estimation of Unconfined Compressive Strength by Spatial Interpolation Using Non-Geostatistical Methods and Artificial Neural Networks

Thongchai Phothong a, and Boonsap Witchayangkoon

a*



A B S T R A C TArticle history: Received 22 September 2014 Received in revised form December 19, 2014 Accepted December 24, 2014 Available online January 05, 2015 Keywords: soil engineering properties; soil testing; ANN.

This study applies spatial interpolation to estimate soil engineering properties by using previous information in the neighborhood areas. This study focuses on soft clayey Bangkok soil data in the Bangkok Thailand. The non-geostatistic and artificial neural networks (ANN) methods are compared to estimate unconfined compressive strength of soil. The non-geostatistics are inverse distance weighted, triangulation, natural neighbor, b-spline approximation, cubic spline approximation, global thin plate spline, local thin plate spline and thin plate spline. For this study, ANN is the four layers feed forward neural networks with error back-propagation learning. From the computation with the testing data, the cubic spline approximation gives the lowest RMSE. ANN is also applicable with more input data.


1. Introduction

Any construction projects, it is important to know engineering soil properties in the construction area. For both feasibility and construction phases, knowing accurate engineering soil properties require field sample collections via borings which are somewhat costly and time consuming. By this, if one has estimated information of engineering soil properties, it would be of great benefit. Also, the estimated information can be used to crosscheck the real data. By having database engineering soil properties of previous multiple testing, this work therefore tries to apply and test various methods to see feasibility in estimation of engineering soil properties.


50 Thongchai Phothong, and Boonsap Witchayangkoon

2. LiteratureReview

Suwanwiwattana et al. (2001) constructed geotechnical database system and subsoil model interpretation of Bangkok Clay, Thailand. The SPT soil engineering property was used to represent soil consistency that was interpolated by GRASS's spline module. The accuracy of the model was tested by manually interpreted and the comparison was satisfactory.

Soralump et al. (2010) developed a soil database system for infrastructure development. The soft Bangkok clay was used to be the case study same as Suwanwiwattana et al. (2001). Soralump et al.(2010) was emphasis on a support soil data system. It is not a soil property estimation system. But in the system has an evaluation algorithm for variation of soil properties with in 5x5 square kilometers.

For soil engineering estimation, Al-Ani et al.(2014) used 8 interpolation techniques, IDW by geostatistical analysis, Diffusion, Global Polynomial, Kernel, Ordinary Kriging, Universal Kriging, Spline and IDW to estimate standard penetration test value in Surfers Paradise, Australia. All methods were run and evaluated by observed data with the same position. The IDW by spatial analysis with parameters 2.719e-05 for output cell size, power by 2, search radius fixed and distance 0.25 km was outperformed other methods.

Gangopadhyay et al. (1999) illustrated a powerful performance of a combination tool of ANN and GIS. The ANN used to classify subsurface aquifer characteristics and GIS received that data to estimate depth-averaged aquifer parameters such as transmissivity, leakage factor and storage coefficient. The multilayer perceptron with the back-propagation algorithm was used. The input were location (x, y), depth, z, and extend of particular type material type, z-from and z-to. The output information was the aquifer material present for the input depth zone. The samples were divided to four strata by the variation of sand frequency, each strata was 50 meters.

Zhao et al. (2009) used the ANN to predict high resolution of soil texture map because though field survey is time consuming and expensive. The input of the ANN were coarse resolution and DEM data. The coarse composes of clay map, sand map. The DEM data is soil terrain factor map, soil drainage map, soil deliver ratio map and vertical position map. The relative overall accuracy was 88% for clay content and 81% for sand content.

Some main ideas from Jain et al. (1996), the ANN applications can apply to pattern classification, clustering/categorization, function approximation, prediction/forecasting, optimization, content-addressable memory and control. For nonlinear prediction problems, the


51

notable solving networks is feed forward network with two hidden layers combines to supervised learning paradigm and back propagation learning algorithm.

This study examines eight non-geostatistical methods and the artificial neural networks to simulate an engineering soil property (unconfined compressive strength) for Bangkok areas, Thailand. The target of the development is that for a certain situation that the researching results can be used to increase confidence of an engineer while decrease cost of soil investigation.

3. Methodology

Bangkok has 1,568.737 km2 covering coordinates (1491347, 643245) and (1543301, 709475) (N, E) with about 6 million people (SED, 2012). This research used 74 sites with 155 soil boring data to build a database of soil engineering properties. The coordinates of each boring are pinpointed with Goggle Earth. The depths of the soil bore holes vary from 21 to 79.775 meters. This research consider only unconfined compressive strength (Su) because of testing consistency. The Su at depth 9.25m are selected to evaluate the interpolation methods because of number of data. The 86 samples at the selected depth are divided to two sets. The 64 samples for estimating and 22 for testing as shows in Figure 1. The non-geostatistics is shown by Li and Heap (2008) addressing that non-geostatistics is inverse distance weighted, triangulation, natural neighbor, spline approximation and thin plate spline (TPSP). The foundation equation of spatial interpolation at an unknown point by surrounding points is illustrated by weighted averages:

(1),

where ̅ is the estimated data value of the unknown point . is the observed data value at the known point . is the weight of known points. The inverse distance weighting (IDW) method uses weighted as an inverse function of the distance, , from unknown point to known points.

/

∑ / (2),

where is a power parameter set to 2 and is number of sampled points. The triangular irregular network (TRI) method estimates the data value of unobserved point from connected surrounding points that forms a series of triangles. The natural neighbors method (NN) combines the benefit of the nearest neighbors and TIN method. The NN creates a Delauney Triangulation of

z̄ ( x0)=∑i=1

n

λ i z ( xi)


the known points then weights their values by proportionate area. The splines are series of polynomials uses to determine a curve and a surface that represent known points. The b-splines (BSP) are a basic of spline. The spline composes of knots and b-splines. The cubic spline interpolation (CSP) estimates unknown values by a polynomial that continuous though to the second derivative. The thin plate spline (TSP) is surface approximation by splines that try to form curve surfaces. The global thin plate spline (GTSP) is a technic to fit b-spline. The local thin plate spline (LTSP) method is an extension of the TSP. The LTSP uses not more than 10 closest observed points in the estimation processes. The System for Automated Geoscientific Analyses (SAGA) software is used in this study that provides eight non-geostatistics inverse distance weighted, triangulation, natural neighbor, b-spline approximation, cubic spline approximation, global thin plate spline, local thin plate spline and thin plate spline (TIN).

(a) (b)

Figure 1: The ANN model (a) model 1 (b) model 2.

For The Artificial Neural Networks has two model both model are feed forward network. The first model has 50 unit for input layer, 50 unit for hidden layer one and two and one unit for output layer. The 50 input units composes of 16+16 difference in north and east direction between unknown point to known point, 16 units of Su value of known point and last 2 input units are north and east coordinate of unknown point. The output unit is Su of known point. The first model shows in Figure 2(a). The second model includes Su data from soil strata 9.00 meter, 47 samples and 10.5 meter, 43 samples. Both are using for training data set.


53

IDW NN

TSP TRI

GTSP LTSP

BSP CSP

Figure 2: The Eight Non-Geostatistic Results.

The input of the ANN model 2 has 194 units. The 16 samples from 9.00meter layer each samples has 4 units, DN, DE, DZ and Su that means 54 units. Same as 9.00 meter layer, the input units from 9.25 and 10.5 meter layer are 54 units. Including 2 units, N and E from unknown point


on 9.25 meter layer. The ANN model 2 shows on Figure 2(b). The number 16 samples from each model is the nearest points around the unknown point, Prasomphan and Mase (2013). The 64 samples for estimating will use to be training data and validating data set because of the limit of samples. The learning in ANN is error back-propagation and activation function is logistic. The different between the observed and calculated values of all estimations are assessed by the root mean squared error (RMSE).

∑ (3),

where is original value subtract estimated value. N is number of points to estimate.

Figure 3: The Testing Values


Figure 2 shows results of non-geostatistic estimations. The b-spline and the local thin plate spline approximation cannot represent a good soil engineering property. The natural neighbor and the triangulation give the same triangle pattern of estimation. The inverse distance weighted, thin plate spline and global thin plate spline gave a smooth estimation but more concentrate on training data set.

The cubic spline and local thin plate spline estimation gave the good distribute and good

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 220

10

20

30

40

50

60

70Testing Data

Observed IDW TRI NN BSP CSP

GTSP LTSP TPSP ANN1 ANN2

Sample Number

Valu

es, k

sc.


55

results. The graph of interpolation results is illustrated on Figure 3. The original testing data is shown on dash line graph. The different of original values and interpolation values are spike on points 1, 3, 7, 11 and 18. All spike points are around the middle of the area. The RMSE is in Table 1. The best is cubic spline spline estimation, 7.42 that relate to the Figure 3. The best ANN model is ANN model 2.

Table 1: The RMSE

For the ANN model, the first model (ANN1) uses only data in the same layer. The RMSE is 11.46 it just similar most of non-geostatistic methods. To increase the number of samples to train the ANN model in this test is use data of different layer with significant improvement.

5. ConclusionandRecommendation

With spatial interpolation technique, this study can interpolation of soil unconfined compressive strength of clayey Bangkok soil data in Thailand. The non-geostatistic and artificial neural networks (ANN) methods are compared to estimate unconfined compressive strength of soil. The non-geostatistics are inverse distance weighted, triangulation, natural neighbor, b-spline approximation, cubic spline approximation, global thin plate spline, local thin plate spline and thin plate spline.

This study found possibility that ANN can be used to predict unconfined strength of engineering soil property. Training the ANN model with high number of data is importance. To increase the number of training data, the samples in different layer can be used. After multiple computational experiments with the testing data, the cubic spline approximation gives the best RMSE while ANN is the second best.

Number Method RMSE1 IDW 11.382 TRI 10.513 NN 10.864 BSP 12.865 CSP 7.426 GTSP 11.257 LTSP 13.818 TPSP 11.289 ANN1 11.4610 ANN2 9.42


6. References

Al-Ani, H., Oh, E., Chai, G. and Al-Uzairy, B., "GIS-Interpolated Geotechnical Zonation Maps in Surfers Paradise, Australia", 2014

Anil K. Jain, Jianchang Mao and K. Mohiuddi, "Artificial neural networks: A tutorial", 1996

Gangopadhyay, S., Gautam, R.T. and Gupta, D.A., "Subsurface Characterization Using Artificial Neural Network and GIS", 1999

Jeff Heaton," Introduction to Neural Networks for Java", 2008

Li, J. and Heap, D. A.," A Review of Spatial Interpolation Methods for v", 2008

Prasomphan, S. and Mase, S., "Generating Prediction Map for Geostatistical Data Based on an Adaptive Neural Network Using only Nearest Neighbors", 2013

Soralump, S., Kunsuwant, B., Mairaing, W. and Surinkum, A., "Development Of Soil Database For Supporting The Development And Maintenance Of Infrastructure: A Case Study Of Soft Bangkok Clay", 2010

SED., 2012 STATISTICAL PROFILE OF BANGKOK METROPOLITAN ADMINISTRATION", Strategy and Evaluation Department of Bangkok Metropolitan Administration, 2012

Suwanwiwattana, P., Chantawarangul, K., Mairaing, W. and Apaphant, P., "The Development of Geotechnical Database of Bangkok Subsoil Using GRASS-GIS", 2001

Thongchai Phothong earned his Bachelor Degree (Civil Engineering) from King Mongkut's University of Technology Thonburi (KMUTT), Thailand, master degree in Geotechnical Engineering also from KMUTT. He is a PhD candidate at Thammasat University Thailand. He is a lecturer at KMUTT. He is interested in spatial technology and applications.

Dr. Boonsap Witchayangkoon is an Associate Professor of Department of Civil Engineering at Thammasat University. He received his B.Eng. from King Mongkut’s University of Technology Thonburi with Honors in 1991. He earned his PhD from University of Maine, USA in Spatial Information Science & Engineering. Dr. Witchayangkoon current interests involve applications of emerging technologies to engineering.

Note: The original of this article has been submitted to 2nd International Workshop on Livable City 2014 (IWLC2014), a Joint Conference with International Conference on Engineering, Innovation, and Technology (EIT), held at Tabung Haji Hotel, Alor Star, Malaysia, during December 9-11, 2014. According to the IWLC2014 Conference Committee, this paper was given Technology Best Paper Award.



Call-for-Papers:





More Research Publication Available at

http://TUENGR.COM/ATEAS

Submit your science research today.

v4(1) 2015:: american transactions on engineering & applied sciences

Documents

usa associate professor

italy associate

italy assistant

south dakota state university

thailand associate editors

mayaguez associate professor

phd associate professor

mayaguez assistant