[american institute of aeronautics and astronautics 9th annual international energy conversion...
TRANSCRIPT
Design and Analysis of a Multi-Window Aperture Structure
for a Small Particle Solar Receiver
Ioana D. Broome* and Fletcher Miller
†
San Diego State University, San Diego, CA 92182
This paper presents the design and analysis of a multi window toroispherical aperture
structure to be used on a Small Particle Heat Exchange Receiver (SPHER). SPHER consists
of a pressure vessel filled with a mixture of sub-micron particles and gas, with optical access
for concentrated sunlight on one end. Solar radiation is concentrated with a field of
heliostats and focused through the multi- window structure inside the vessel where the light
is absorbed volumetrically by the gas particle mixture. The spherical cap structure
presented in this paper serves as the head of the pressure vessel which comprises the
receiver. The interior walls of the receiver are under hydrostatic pressure of up to 5 bar and
the interior surface of the insulation protecting the metal will get as hot as 1000 Celsius. The
structure design has been continuously optimized in efforts to improve the physical and
mechanical properties of the receiver’s windows and increase the structural efficiency of the
solar receiver. Structural efficiency means minimizing the mass while maximizing the design
factor of safety. Subsequent to the studies and analyses ran on different geometries, a convex
stainless steel toroispherical cap under tension was designed, with six concave fused silica
windows mounted on it. The design choice is due to the cost and complexity of fabrication of
a larger, single fused quartz window, as well as maintenance and replacement of parts. The
final dimensions and geometry were established as a result of a series of static and thermal
analyses. It was found that for a minimum calculated factor of safety of 4, the pressure vessel
wall thickness must be at least 1.5 cm with thickened cylindrical section of 2 cm. The
windows minimum thickness was 1 cm.
Keywords: small particle solar receiver, pressure vessel, secondary concentrators, aperture structure
Nomenclature
���� = maximum Von Mises stress (Pa, MPa) H = height (m)
���� = maximum displacement (mm) SS = stainless steel
���� = maximum strain T = temperature (°C)
�� = principal stress 1 (Pa, MPa) FOS = factor of safety
� = principal stress 2 (Pa, MPa) P = pressure (Pa, MPa)
R = radius of circle / arc (m) V = volume (m3)
h = height (m) M = structural mass (kg)
t = thickness (m) = material density (kg/m3)
� = angle (°) Φ = diameter (m)
I. Introduction
ADIATION in the form of light and heat from the sun is what we call solar energy and it is an outstanding
source of renewable energy. Solar energy has been harnessed by people ever since the dawn of time. It is free,
renewable and only a minute percentage of solar heat is being used to help produce electricity. Solar power plants
are becoming more prevalent and they are being perfected all the time with the ultimate goal to produce cheap, non-
* Graduate Student, Aerospace Engineering, 1885 Diamond St, Apt. 320, San Diego, 92109, AIAA member
† Assistant Professor, Dept. of Mechanical engineering, MC 1323, San Diego, CA 92182
R
9th Annual International Energy Conversion Engineering Conference31 July - 03 August 2011, San Diego, California
AIAA 2011-5900
Copyright © 2011 by Ioana D. Broome. Published by the American Institute of Aeronautics and Astronautics, Inc., with permission.
polluting and renewable energy. When developing such a facility there are many factors that have to be considered
regarding placement, weather and climate, structure and optics, cost and maintenance, and plant efficiency. All of
these affect the levelized cost of energy that is used as a metric for solar power plants.
Because of the rising levels of CO2 emissions and other energy concerns, the Solar Thermal Power Plants
(STPP) are becoming important candidates for producing clean, renewable energy4. The low flux density at Earth’s
surface makes it nearly impossible to heat up a Heat Transfer Fluid (HTF) such as air, steam or natural gas, to
temperatures adequate for industrial applications. In order to attain temperatures required to operate a
thermodynamic cycle, a STPP requires optical concentration. Solar radiation is converted to thermal energy inside
the receiver and later to electricity. Among the pioneers of solar energy technologies, the US developed tube
receivers in the late 90’s. Researching a cheaper, simpler and more efficient solution, Europe later developed the
volumetric receiver. The volumetric receiver is more flexible than the tube receivers due to its three-dimensional
configuration and its functionality.
From the structural point of view, the way to increase the plant efficiency is to design a capable and economical
structure that will maximize the intake of solar heat and minimize the thermal losses. Probably the most important
part of the receiver structure is the light and heat collector. The Small Particle Heat Exchange Receiver (SPHER)
that is currently being developed at San Diego State University will collect concentrated solar radiation through
several transparent quartz windows that can withstand the high pressure and high temperature located inside the
receiver. The windows are essential to this design to allow for volumetric absorption, and secondary concentrators
are proposed to minimize the size of the windows needed. The concept of volumetric absorption is explained in
Section II.
Figure 1. Small Particle Heat Exchanger Receiver (SPHER) Placement
In order to build the most efficient and cost-effective pressurized air-heating solar collector, a literature review
and study of similar structures was performed. The very few previous projects that have developed similar products
use windows shaped as shells of revolution that withstand the operating conditions of the receiver, including high
temperature and pressure. The first part of the study focused on these previously developed structures as a starting
point for improving this design and building the SPHER.
In this research, flat and curved profiles for the pressure vessel head were compared and a weight minimization
study was used to choose the depth of the head. Next, the profile of the pressure vessel head was analyzed and the
curvature reshaped as a result of the stress analysis performed on it. Finally, the vessel body was structurally
optimized and a series of stiffeners are evaluated. After determining the ideal thickness for both assembly parts:
head and body, respectively, a final design was presented. After the final design was confirmed, a material selection
step was carried out, choosing the material with the highest performance under the given operating conditions.
II. Literature Review and Research of Similar Products
A. The SPHER
The small particle solar receiver is a design introduced by A.J. Hunt in 1979.1 It is basically a volumetric
pressurized solar receiver that uses very small particle clusters of carbon to absorb the concentrated sunlight and
transfer it to a gas which flows into a turbine where the thermal energy is transformed into electricity. The choice of
carbon is due to its physical, chemical and optical properties.1
Heat and radiation from the sun is reflected from the
heliostats into the receiver where it is trapped. Very small carbon particles are injected into the receiver where they
capture the heat. The extremely hot mixture of air and carbon particles exits the receiver and enters a gas turbine
which powers a generator to generate electricity. The carbon particles actually burn up before they exit the receiver
so the hot mixture is mostly air.
The previously tested SPHER was limited to atmospheric operation and 30 kW of power. A variation of this
small particle solar receiver is currently being developed at San Diego State University (SDSU).10,14
Concentrated
sunlight will enter the pressurized vessel through a series of fused silica windows shaped so that they can withstand
a maximum hydrostatic pressure of 5-10 bar (0.5 – 1 MPa).2,10
The gas temperatures inside the receiver will reach
1000° Celsius. The pressure vessel’s inner walls are insulated as well as the head structure that contains the
windows. A conceptual schematic showing a single window is presented in Figure 2
Figure 2. SPHER schematic with a single window
The solar radiation (1) enters the receiver through the window or window assembly (2). Gas particle inlets (3)
inject a mixture of air from the turbine’s compressor (8) and small carbon clusters (4). The mixture circulates up to
the window area, where the gas temperature reaches the hottest point (5) and absorbs the heat. The hot mixture of air
and carbon clusters is transported through an outlet (6) and powers a gas turbine (7). The mechanical power is
transferred to a generator (9) which produces electricity.3
B. Volumetric Receiver
The basic operating principles of a volumetric receiver are: 4
• A cluster of porous shapes, foam, metal, ceramic or other
adequate materials with a specific porosity are placed in a
volume inside the solar receiver so that the solar radiation
is absorbed in the depth of the structure
• The porous material inside the receiver is heated by the
concentrated solar radiation. The working fluid passes
through the volume and is heated by forced convection
transforming the radiation into thermal energy.
The absorber can be ceramic or metallic. Ceramic absorbers can
reach higher temperatures than metallic ones, but are prone to
cracking.
Figure 3 Volumetric Receiver Schematic
Volumetric receivers have been studied in the past 30 years and classified in 4 subgroups: 4 Phoebus-TSA, SOLAIR,
REFOS and DIAPR. The classification above was made based on the air pressure and materials
• Phoebus – TSA type: open-loop volumetric receiver with metallic absorber
Projects of this type: MK-1, Sulzer 1 (MK-2), Sulzer 2, Catrec 1, TSA, Bechtel 1, Bechtel 2, Catrec 2,
SIREC
• SOLAIR type: open – loop volumetric receiver with ceramic absorber
Projects of this type: SANDIA foam, CeramTec, Conphoebus-Naples, Selective Receiver, HitRec 1,
HitRec 2, SOLAIR 200, SOLAIR 3000
• REFOS5and DIAPR
6 type: closed- loop volumetric receiver with metallic or ceramic absorber
Projects of this type: PLVCR-5, PLVCR-500, DIAPR 30-50, DIAPR multistage, REFOS, SOLGATE
C. REFOS (Solar-Hybrid Gas Turbine-based Power Tower Systems)
In 1996, the German Aerospace Center called Deutsches Zentrum für Luft-und Raumfahrt (DLR) developed a
pressurized volumetric solar receiver as part of the REFOS project, tested at Almeria, Spain. The receiver system
consisted of a module of 350 kW nominal power, which included a pressurized volumetric receiver and a secondary
concentrator. The system was operated for 90 hours and it reached a maximum power of over 400 kW. There was
only a minor mirror gluing issue; however the test was able to demonstrate the technical feasibility of the modular
receiver technology.5 The REFOS module operating conditions are: 350 kW of thermal power, 15 bar operating
pressure, up to 800° C air outlet temperature and a receiver efficiency of 80%. The schematic of the receiver as well
as an image of the secondary concentrator is shown in Figure 3.5
Figure 4. Schematic of the solar receiver with the secondary concentrator by DLR (left) and the DLR
secondary concentrator (right).5
The receiver window has a semi-elliptical profile with a 620 mm diameter at the open end and a 420 mm depth.
Its thickness is 8 mm and it has been certified up to a 19.5 bar pressure. The Compound Parabolic Concentrator
(CPC) has a hexagonal entry aperture with an inscribed circle of 1.2 m diameter. Its acceptance angle is 20° and its
length is 1m. The flat trapezoidal aluminum plates that form the structure are water-cooled and have backsilvered
glass mirrors glued onto the plates.5
D. DIAPR (Direct Irradiated Annular Pressurized Receiver)
This type of receiver is a volumetric (directly irradiated) cavity receiver. Its operating pressure is between 10 and
30 bar, the outlet gas temperature is up to 1300 °C and its aperture radiation flux is 10 MW/m2. The DIAPR receiver
can be used at a power of 100 kW to power a Brayton cycle. The three main innovative components of the DIAPR
receiver are: the high temperature ceramic Porcupine absorber, the fused silica Frustum-Like High–Pressure (FLHP)
window and the two stage secondary concentrator followed by the KohinOr light extractor.6
Figure 5. Schematic cross-section of the DIAPR receiver (left) and window (right).6
The Frustum-Like High-Pressure window’s purpose is to separate the receiver cavity from the outside medium
and allow operation at high pressure while minimizing the reflection losses of solar radiation which enter the
receiver. A closed- loop test using carbon dioxide (CO2) as the working fluid showed that the window survived a
pressure of up to 30 bar (3 MPa), contamination from dirt and large amounts of graphite from the CO2 deposited
onto its surface. Ray tracing calculations showed that the reflection losses of the window are about 1%. Window
dimensions (in millimeters) are shown in Figure 5. 6
E. Scope of the research (design and analysis of structure)
The work done in this research builds on the effort started over 30 years ago at the Lawrence Berkley Laboratory
and it supports a larger project with the purpose to help develop and test a prototype small particle solar receiver like
the one presented in Section II -A. Its relevance to the big project is to calculate the structure that will withstand the
workloads as well as show its structural capabilities as part of the whole power plant assembly.
The Small Particle Heat Exchange Receiver (SPHER) currently in development at SDSU proposes to use a
suspension of small carbon particles entrained in the gas as the absorber, in contrast to previous absorbers described
above that are fixed structures. The pressure vessel is much larger than anything developed before for this purpose,
and it has multiple large windows mounted on one end. The work of this research focuses on the structure of the
receiver, with special attention given to its head assembly that includes the windows and their mounting mechanism,
as well as the secondary concentrators. The secondary concentrators will reflect more of the light inside the vessel
and more solar radiation will be absorbed by the small carbon particle clusters. No optical analysis is performed on
the secondary concentrators.
Figure 6. SPHER assembly showing windows and secondary concentrators – artist’s rendition only
III. Conceptual Design
The receiver initial dimensions chosen for analysis were 3 meters in diameter and 5 meters tall. 3 The multiple
windows concept is investigated due to possible ease of manufacturing as well as lowering the costs compared to a
single large window. An array of 6 windows is proposed to be installed on the end oriented towards the heliostats.
They are about 1 meter in diameter, having the same profile as the head of the pressure vessel, but instead of being
in tension they will be compressed by the pressurized gas. 7
A. Pressure Vessel
Before the preliminary design phase started, the failure criteria were selectively analyzed. The failure criterion
was chosen to be Von Mises. This stress/failure theory is favorable for ductile materials. Although it is less
conservative than the Maximum Shear Stress Theory (Tresca yield criterion), it is more conservative than the
Maximum Normal Stress Theory (Rankine criterion). A comparison graph of all three theories is shown in the figure
below. 8
Legend
DE = Distortion Energy Stress Theory (Von Mises)
MN = Maximum Normal Stress Theory (Rankine)
MS = Maximum Shear Stress Theory (Tresca)
F
Figure 7. Projection of the Failure Criteria on
the σ1, σ2 plane. 8
1. Head Structure Profile
The first stage of the conceptual design phase was calculating and choosing the shape profile of the pressure
vessel head. The first step was to figure out whether the design should be a flat, convex‡ or concave
§ profile for the
spherical cap. Intuitively it can be said that the flat head would be very inefficient because of the large deformations
it would undergo or very costly because it would have to be very thick to withstand the constant pressure of 5 bar.
However, a flat head might provide some advantages for mounting the secondary mirrors so all three designs were
analyzed. The dimensions are listed in Table 1Table 1. For this analysis the rest of the pressure vessel did not need
to be modeled, the head shapes being enough to provide a good comparison. The caps (convex and concave) are
spherical caps cut at a 60° angle for reasons presented in the angle study section (III,A,2). The static analysis was
performed at room temperature and a pressure of 5 bar (0.5 MPa) was applied.
Table 1. Analysis results of the three head geometries
Analysis results Pressure only (5 bar) Pressure (5 bar) and temperature (100 °C) units
Flat Convex cap Concave cap Flat Convex cap Concave cap
� �� 738 30.501 30.584 920 459.7 420.9 MPa
� �� 46.87 0.1277 0.13 57.25 0.967 0.714 mm
� �� 2.671e-3 9.872e-5 9.935e-5 2.635e-3 1.138e-3 1.003e-3
The thickness chosen for this study was 1.54 cm (~1 in). The mass of the curved ones was 1887.17 kg, while the
mass of the flat head was, 23.89% lower (1436.34 kg)
‡ Convex spherical cap: has the pressure applied inside its curvature and it is in tension
§ Concave spherical cap: has the pressure applied on the outside surface and is under compression
Figure 8 Geometry of the three concepts from left to right: Flat, Convex, Concave
As it can be deducted from the results in Table 1, the convex head analysis returned better results (lower stress
and maximum deformation) and was therefore chosen to be the design solution for the pressure vessel head. The
pressure was applied on the inside curved surface.
2. Angle Study7
A weight minimization study (see Figure 9) was carried out next, meaning the least weight and therefore most
cost efficient profile of the spherical cap was calculated, which was represented by a shell of revolution of height H,
radius R, angle θ and thickness t (See Figure 10a). Using the thin shells of revolution theory9, the minimum
thickness t was calculated as described in Ref.7 for the optimum angle. Shells of revolutions are shells that have the
same exact cross-section at any plane that contains the axis of rotation we would use to intersect the shell (e.g. a
shape made from a profile sketch revolved around the axis of rotation)
Figure 9. Weight minimization study; θ=5° to 90°
As Figure 9 shows, the most efficient geometry for the head, corresponding to the minimum mass with the
maximum strength was the spherical cap cut at an angle � � 60°.
0
50
100
150
200
250
300
350
400
450
500
5 10 15 20 25 30 35 40 45 50 55 60 65 70 75 80 85 90
Mass
(k
g)
Angle (deg)
5 bar
10 bar
15 bar
20 bar
25 bar
Mimimum
Mass
Figure 10. Initial head profile (a. left) and optimized head profile (b. right). Dimensions in meters
3. Head curvature update
First, the calculated 60° profile was selected and a set of static analyses was performed on the head shaped as
seen in Figure 1010a. Stress concentrations were found around the joints (where the heads weld onto the cylindrical
body). As a result of these high stress concentrations, the ends are chosen to be filleted, thus creating a toroispherical
head. The toroispherical head simulates the ellipsoidal one by a compound curve made of a crown radius and a
knuckle radius (see Figure 10b). The knuckle radius has to be big enough to minimize the hoop stresses around this
region. According to the ASME Pressure Vessel Code, the knuckle radius (R1) has to be at least 6% of the crown
radius (R 1.73).11
In this case it is well over the minimum required in an effort to minimize the hoop stresses.
4. Pressure Vessel Bottom Profile
The design of a pressure vessel, especially one subjected to a thermal load, is far from simple and therefore a
series of steps need to be performed. According to the ASME Boiler and Pressure Vessel Code, section VIII, the
design factor of safety for a pressure vessel has to be a minimum of 2.5 – 4.11
Based on the head curvature study7
and the fact that the pressure vessel in question is not a storage tank, the flat bottom was eliminated in the
conceptual design phase. Two profiles were modified and analyzed. No thermal effects were considered at this stage
and the pressure applied was 5 bar (0.5 MPa). The material applied was stainless steel type 316, known for its high
temperature strength.
To make sure that the vessel will withstand the pressure inside without excessive deformation, a series of static
tests were performed. As a design outcome of this test, the geometry of the vessel was also improved, to
accommodate the pressurized gas inside the receiver. The design parameters, such as the bottom shape and the wall
thickness, were updated until the overall design was within the ASME constraints. The outcome and results are
presented in Table 2.
Table 2. Pressure Vessel Bottom Profile Selection
Profile 1 2 3
Thickness 2.5 cm 5 cm 2.5 cm
Shape Concave (compression) Concave (compression) Convex (in tension)
Min FOS 0.936 3.036 11.339
Stress
plot
Starting from a typical pressurized vessel, such as a soda can or a champagne bottle, the initial stage of the shape
selection is modeled accordingly. A bottom under compression (concave) is given to the pressure vessel and a
uniform thickness is applied. When the 2.5 cm initial selected thickness fails the static test, the thickness (and with it
the mass) is doubled. The factor of safety for the thicker vessel (5 cm) is still below the minimum accepted. In an
effort to minimize the mass, the convex (in tension) bottom is analyzed. A uniform thickness of 2.5 cm is given and
the minimum calculated factor of safety increased a great deal. Much more satisfying, the third profile was selected
for further analyses. During this design phase, the calculated factor of safety was brought from 0.936 to 11.339,
another proof that smart design can save unnecessary use of extra material.
5. Stiffeners
As a result of the pressure vessel bottom profile section, the convex bottom was selected as effective. The high
stress area for this design is around the cylindrical portion of the vessel. To help minimize or eliminate hoop stresses
on the cylindrical part without thickening the wall everywhere structural stiffeners are proposed. Three types of
stiffeners were analyzed.
Table 3. The three types of stiffeners tested
Stiffener One ring Three rings Thicken cylindrical part
Mass (kg) 392.96 1178.88 2416.8
Min FoS 11.36 11.388 17.148
Section
stress
plot
Assuming weight is not a factor but structure quality alone, the third proposed stiffener type was selected for
further studies. In Section IV an optimization of the wall thickness is performed and thus, the weight of this model
will be minimized.
6. Pressure vessel head with portholes analysis
A final shape was calculated for the pressure vessel but in order to install windows on one head, portholes need
to be created on the respective head. Holes will generally weaken the head which provides structural support for the
windows and the windows are desired to be as large as possible so a compromise needs to be made. The array was
established to be one central window port and five of the same size ports equally spaced around the middle one.
A static analysis was performed on the steel pressure vessel head with 6 portholes cut outs. This is called the
aperture structure. The inner pressure difference applied to the stainless steel 316 structure was 0.5 MPa (5 bar) and
the operating temperature was 25 °C. Circular loads corresponding to the windows were applied around each
porthole. Figure 11 illustrates the Von Mises stress field on the top (a) and the bottom (b) of the head. The stress
plot reveals stress concentrations around the window portholes which calls for some structural reinforcements
around the edges. Special flanges welded onto the edges could possibly solve this problem. Figure 12 is a
displacement plot of the head shown on the top (a) and bottom (b) of the same structure, the head of the pressure
vessel. The displacement plot exposes a radial deformation field around the center porthole. This type of
deformation cannot be avoided in the convex shell of revolution head geometry, so the solution proposed is to leave
enough clearance between the edges of the secondary concentrators that are to be installed on top of this structure.
The clearance is necessary to avoid surface interference and cracking of the secondary concentrators.
Figure 11. Stress plot on the: (a. left) top and (b. right) bottom of the head structure; Pressure = 0.5 MPa @
room temperature
Notice how the stress concentrations are seen only on the inside surface of the head structure due to the hydrostatic
pressure. The stresses are minimized and/or dissipated by welding of the mounting special flanges onto their edges.
Figure 12. Displacement plot on the: (a. left) top and (b. right) bottom of head structure; Pressure = 0.5 Mpa
@ room temperature
B. Window
The window is one of the most difficult parts of the assembly to fabricate in a custom shape and install on a
stainless steel structure. The selected material is fused silica due to its extremely low thermal expansion and very
good optical properties. It was showed by a weight minimization study10
that the optimum cut angle for the spherical
cap representing the window under compression is 60°. The portholes created on the head aperture structure exposed
to the concentrated solar radiation are no larger than 1 meter in diameter due to the need of structure for support as
well as to prevent intersection of the windows and/or flanges due to head curvature and thermal expansion (see
Figure 14a). Special flanges welded to the portholes are used for windows installation. Thus, the window outer
diameter must be less than or equal to 0.96 meters. A surface with a 0.96 meters diameter was revolved and a design
study was performed to calculate the optimum thickness. The tensile strength of fused silica is 50 MPa and a
minimum factor of safety of 5 was set for the window’s principal stresses. This means the maximum acceptable
principal stress can be no higher than 10 MPa. An optimization study was carried out following the constraints
mentioned earlier and thus the minimum allowable window thickness was found to be 1 cm. Because the edges are
especially problematic when it comes to principal stresses, a shape optimization step was performed changing the
curvature of the window from spherical cap to toroispherical.7
Table 4. Window curvature study; from the spherical cap with edges at an angle (profile 1) to the
toroispherical head with straight edges (profile 3)
Profile 1 Profile 2 Profile 3
Sketch
Max
principal
stress
(MPa) 7.053 3.78 1.445
Mass (kg) 20.86 24.36 28.98
All the results were within the principal stress limits established (max 10 MPa) and therefore they were all
feasible designs. As a result of this curvature study, the potential client is given the option to choose between cutting
fabrication costs by saving weight or maximizing performance where weight is less important than performance. For
future assemblies and analyses, the third profile is used. For more info on the window assembly and seal see Ref. 7
and Ref.10.
C. Secondary concentrators
The pressure vessel with windows is directly exposed to concentrated solar radiation coming from the heliostat
field. The addition of some secondary concentrators is necessary because they need to redirect radiation from
impinging on the solid portion of the head (see Figure 14a) into the windows. This will prevent overheating the
stainless steel head and allow all the light to be captured. The pattern is chosen to be a pentagonal aperture in the
center surrounded by hexagonal aperture concentrators. The “soccer ball” model (see Figure 14b) is chosen for best
fit. Their pentagonal aperture has an inscribed circle with the diameter Φ1.652 meters and the hexagonal apertures
have an inscribed circle with a Φ1.905 meters diameter.7 The acceptance angle is 132.12° at this stage in the design
but this could be changed once the heliostat field is chosen. This is shown in Figure 13 below.
Figure 13. Field concentration angle and secondary assembly acceptance angle
Figure 14. Solar receiver with windows exposed to concentrated radiation (a. left) and “soccer ball” pattern
secondary concentrators assembly without windows (b. right)
Backsilvered mirrors or a high temperature aluminum alloy are recommended for the fabrication of secondary
concentrators. A simple static analysis was run on the concentrators’ structure to assure its ability to withstand the
pressure from inside the vessel. A maximum displacement of 0.42 mm was exhibited by the edges around the center
concentrator and a maximum Von Mises stress of 41.174 MPa was calculated around the special flanges when the
concentrators were installed onto the aperture structure. This means a 66.54% decrease in maximum stress due to the
flanges and installed concentrators. The plots are shown in Figure 15.
Figure 15. Stress (left) and displacement (right) plots on the head assembly with concentrators and flanges.
IV. Design Optimization and Detailed Model Analysis
A. Material selection
Due to its high chromium content, very good corrosion and oxidation resistance and surface quality, the
austenitic class of stainless steels is selected for the high-pressure high-temperature pressure vessel head.12
A few
examples of steel belonging to this class are types: 201, 301, 304, 310,316, 316L, 321 steel, named by the SAE
standard. Four grades of austenitic stainless steel were selected for the pressure vessel design and analysis: SS310,
SS316, SS321 and SS347. Their mechanical and physical properties are listed in the Appendix. They are considered
high alloys due to their high chromium content. The high alloy steels typically exhibit excellent strength at elevated
temperatures along with creep deformation resistance and other environmental attacks such as oxidation and
corrosion.12
Figure 16 shows the effects of temperature on the yield strength of all four materials selected.
Because of all the challenges of fabricating such a complex structure, the actual design can help a great deal with
the structural workloads. A smart material selection is very important and can save future maintenance as well as
minimizes the weight of the structure.
Figure 16. Temperature Effect of Yield Strength of Stainless Steel **
B. Structural Optimization
In order to achieve the desired performance while designing the structure within the limits set by the ASME
standard, there is no need to over-design a bulky vessel, but rather optimize its design variables, such as the
thickness. After calculating the final geometry for the pressure vessel, a structural optimization step was performed,
in an effort to minimize the weight of the product. The optimization study was performed using SolidWorks® and
its optimization package, Design Study.
The design optimization study analyzed different thicknesses ranging between 1 cm and 3 cm for both parts of
the pressure vessel. The step chosen is 0.5 cm and therefore 16 design scenarios were described. The minimum
design factor of safety was set to 4, which limited the maximum Von Mises stress anywhere in the structure to less
than or equal to 77.5 MPa. The goal was to minimize the mass. It was found that for a steady hydrostatic pressure of
up to 5 bar and inside temperatures of up to 800 °C, the minimum required thicknesses of the pressure vessel with
the earlier mentioned dimensions are: 1.5 cm for the head and 2 cm for the middle section. It is desired that the wall
temperature be much lower than 800 °C, but this study was meant to show that the designed pressure vessel would
withstand the maximum operating temperature of stainless steel if it had to.
**
Data from http://www.sandmeyersteel.com/300-series-heat-resistant.html
0
50
100
150
200
250
300
350
0 200 400 600 800 1000
Yie
ld S
tren
gth
(M
Pa
)
Temperature (C)
310S
316
321
347
The optimization study was done for the targeted work conditions, but it was proven that the change in pressure
ratio of the cycle can lead to a greater expansion ratio through the turbine and increase the power output.13
To assure
the correct functioning during off-design operation, the margin of error established for the pressure will be 60%,
which means that the design factor of safety for an 8 bar pressure has to be higher than 4. A new study showed that
the structure could withstand a hydrostatic pressure up to 0.8 MPa (8 bar) but the temperature increase made the
structure fail.
In conclusion, it is advised that the steel inner walls not to reach temperatures higher than 800 °C at any given
time and the maximum allowed pressure at elevated temperatures should be no higher than 0.8 MPa (8 bar). There
are stainless steel types that could operate at temperatures up to 1000 °C, such as 310S, but when pressure comes in
play an insulation jacket is recommended to avoid thermal
cycling, fatigue and ultimately structural failure. For the
complete study see Ref.7.
C. Structural Characteristics
The stainless steel pressure vessel studied in this thesis will
be fabricated out of austenitic stainless steel of type 310S, 316,
321 or 347. Type 310S is preferred due to its higher maximum
service temperature. It is 4.76 meters tall, has an inner diameter
of 3 meters and non-uniform wall thickness of: 1.5 cm for the
toroispherical heads and 2 cm for the cylindrical section,
respectively. The design drawing with dimensions in meters is
shown in Figure 17. The structural capabilities of this structure
are: a constant hydrostatic pressure of 5 bar (0.5 MPa) with the
maximum no higher than 8 bar (0.8 MPa) and a maximum
continuous temperature of 800 °C.7 Cyclic loading, fatigue and
creep were not considered in this study. A future, final design
should include these considerations.
V. Conclusion
Considering factors like lawful constraints, such as the ASME Boiler and Pressure Vessel Code, service limits
and design boundaries, an optimum model for a SPHER receiver was calculated in an effort to minimize the weight
and cost of the material while maximizing performance. The material choice was an important step as it can save the
manufacturer extra work and the financer extra costs. The windows and the pressure vessel heads are loaded on
opposite surfaces and are made of very different materials, fused silica and stainless steel, respectively. Despite
these facts, it was found that the optimum shape profile for both components was toroispherical and the theory of
thin shells of revolutions was used in both cases to help calculate the weight minimization studies. Due to
continuous exposure to solar radiation as well as scattered light a mirrored secondary concentrator was attached to
each individual window. The structural capabilities of the SPHER assembly were calculated as well. Thus, the
maximum allowed continuous temperature inside the vessel was 800 °C and the maximum admissible pressure was
0.8 MPa (8 bar). Since the working gas inside the receiver is targeted to heat up to 1000 °C, the solar receiver
structure must be insulated effectively so that the inner walls will not overheat above 800 °C, to prevent structural
failure. A rendition of the complete exploded view of SPHER assembly is displayed in Figure 18.
Figure 17. Final Pressure Vessel Design;
dimensions in meters
Figure 18. Exploded view of solar receiver
VI. Appendix
Table 5. Material properties in SI units
Property Metric units 310S 316 321 347
Young's Modulus Pa (N/m2) 2.04E+11 2.05E+11 2.05E+11 2.04E+11
Shear Modulus Pa 8.1E+10 8.2E+10 8.2E+10 8.1E+10
Density kg/m3 8040 8070 8050 7970
Poisson's ratio - 0.275 0.275 0.275 0.275
Tensile strength Pa 6.2E+8 6.2E+8 6.6E+8 6E+8
Yield strength Pa 3.1E+8 3.1E+8 2.41E+8 2.8E+8
Specific heat J/kg/K 530 530 530 530
Max service Temperature K 1420 1200 1200 1200
Max service Temperature C 1146.85 926.85 926.85 926.85
Fatigue Strength (107 cycles) Pa 2.28E+8 2.52E+8 2.47E+8 2.52E+8
Thermal conductivity (100 C) W/(m*K) 13.8 14.6 16.3 16.3
Thermal conductivity (500 C) W/(m*K) 18.7 20.942 21.4 21.4
CTE (0-100 C) /K 1.440E-05 1.602E-05 1.674E-05 1.674E-05
CTE (0-200 C) /K 1.620E-05 1.620E-05 1.710E-05 1.710E-05
CTE (0-300 C) /K 1.692E-05 1.746E-05 1.854E-05 1.854E-05
CTE (0-400 C) /K 1.746E-05 1.854E-05 1.926E-05 1.908E-05
CTE (0-500 C) /K 1.908E-05 1.998E-05 2.016E-05 1.998E-05
Max temp w/o
excessive scaling K 1366.4833 1172.0389 1172.0389 1172.0389
Stress relieving temp K 1310-1422 1310-1422 1310-1423 1310-1424
VII. Acknowledgments
I would like to thank Dr. Fletcher Miller, my thesis adviser, for guiding me in the right direction and not letting
me get distracted.
Thanks to Dr. A.J. Hunt, the inventor of the SPHER, for his insight on different design matters.
Thanks to my lab peers for making the work environment pleasant.
Lastly, thanks to Google.org RE<C program for funding this project.
VIII. References 1Hunt, A. J., “A New Solar Thermal Receiver Utilizing Small Particles,” in Proc. Int. Solar Energy Society Conf.,
Atlanta, GA, 1979 2Kitzmiller, K., and Miller, F., “Thermodynamic cycles for a small particle heat exchange receiver used in
concentrating solar power plants”, Journal of Solar Energy Engineering, accepted for publication August, 2011. 3Ruther, S.J., Radiation heat transfer simulation of a small particle solar receiver using the Monte Carlo method,
M.S. Thesis, Mechanical Engineering, San Diego State University, CA, 2010 4Avila-Marin, A.L., Volumetric Receivers in Solar Thermal Power Plants with Central Receiver System Technology:
A Review, CIEMAT, Departamento de Energia, Avda Complutense 22, E-28040 Madrid, Spain††
5Buck, R., Abele, M., Kunberger, J., and Denk, T., “Receiver for Solar-Hybrid Gas Turbine and Combyned Cycle
Systems”, 9th
International Symposium ‘Solar Thermal Concentrating Technologies’ Font Romeu – Odeillo, France,
June 22-26, 1998 6Karni, J., Kribus, A., Doron, P., Rubin, R., Fiterman, A., Sagie, D., “The DIAPR: A High-Pressure High-
Temperature Solar Receiver”, ASME Journal of Solar Energy Engineering, vol. 119, Feb 1997 7Broome, I., “Design and Analysis of a Multi-Window Aperture Structure for a Small Particle Solar Receiver”, M.S.
Thesis, Department of Aerospace Engineering, San Diego State University, CA, December, 2011 (unpublished) 8Moss, D.R., Pressure Vessel Design Manual, 3
rd ed., Elsevier Inc., 2004
9Ugural, A.C., Stresses in Plates and Shells, 2
nd ed. McGraw-Hill, 1999
10Mande, O., “Window and Seal Design for a Small Particle Solar Receiver”, M.S. Thesis, Department of
Mechanical Engineering, San Diego State University, CA, August, 2011 (unpublished) 11
ASME Boiler & Pressure Vessel Code, Section VIII- Rules for construction of Pressure Vessels 12
Oberg, E., Jones, F.D., Horton, H.L., Ryffel, H.H. Machinery Handbook, 25th ed., Industrial Press Inc, New York,
1996 13
Kitzmiller, K. and Miller, F., Effect of Variable Guide Vanes and Natural Gas Hybridization for Accommodating
Fluctuations in Solar Input to a Gas Turbine, ASME 2011 Turbo Expo, June 6-10, 2011 Vancouver, Canada 14
SolidWorks Education Edition 2010-2011, Software Package, SP5.0, Dassault Systemes, Paris, France, 1995-2010 15
Microsoft Office, Software Package, Ver. 2003, Microsoft, Redmond, WA
††
Found at www.elsevier.com/locate/solener