2012-05-02 the science of solvent welding -...
TRANSCRIPT
455 W. Victoria Street, Compton, CA 90220 U.S.A. • www.weldon.com
THE SCIENCE OF SOLVENT WELDING
A technical presentation for specifiers and owners
about the inner workings of solvent cements
By Fabio Castellani
May 2, 2012
Often tim
piping bus
‘reputatio
cementin
the most
joining vin
Let me sta
What this
(primer if
variations
‘magnitud
industrial
pumps, ta
work beca
not used c
When any
made to b
include in
and/or ce
I write thi
needed fo
what is th
results th
e, things that
siness this in
on’. Sometime
g’ (SC) and th
cost effective
nyl and ABS p
art by showin
s pie chart sho
needed) in c
s of the show
de’ of the per
application it
anks, actuato
ause the wro
correctly the
yone argues t
bring the chea
n the specifica
ertified by the
is paper to br
or it to work a
he correct way
at can be exp
t are perceive
turns brings
es this applie
hat is a shame
e, most reliab
piping system
ng something
ows are the c
onstruction (
n cost allocat
rcentage of th
t could be as
rs, etc.). Yet,
ng products a
remaining 99
that all cemen
apest contrac
ation that the
e SC provider
ring clarity to
as advertised,
y to use the p
pected when a
ed to be simp
problems, fai
s to solvent w
e because wh
ble, most resil
s.
that amazes
cost allocation
commercial o
tions dependi
he cement co
high as 1% be
if that minisc
are used, bec
9+% is in jeop
nts are the sa
ctor regardles
e contractor m
(certification
what solvent
, what is the
product, why
all this comes
1
ple or easy are
lures, high m
welding or, as
en properly u
lient, stronge
me every tim
ns for the pur
or residential
ing on the bu
mpared to th
ecause there
cule percenta
ause inferior
pardy.
ame and price
ss of experien
must either sh
that Weld‐O
t welding real
relevance of
it must be us
s together.
e the ones mo
maintenance c
s it is more co
understood a
est and longe
me I look at it
rchase of pipe
). Although t
uilding or mat
he PVF will no
tends to be a
age of the bill
quality is cho
e is the only d
nce or expert
how certificat
On provides fo
lly is‐‐how it w
pipe and fitti
sed in a very s
ost overlooke
costs and with
ommonly refe
and implemen
est lasting way
:
e, valves, fitti
there can cert
terial used or
ot very by mu
a lot more joi
l of material c
osen or becau
differentiator
ize, when Eng
tion or needs
or free), I am
works, why it
ng dimension
specific way a
ed and in the
h all that com
erred to, ‘solv
nted it is bar n
y of permane
ng and ceme
tainly be
sizes, the
ch. In an
ints (but also
cost does not
use the produ
, when choice
gineers refus
s to be trained
frankly stunn
t works, what
ns and toleran
and what are
mes a
vent
none
ently
nt
t
uct is
es are
e to
d
ned.
t is
nces,
e the
Just one n
message i
understoo
connect P
PART 1
As a manu
the differe
the two fa
chemistry
substrate
Before we
but essen
convince t
the Contr
A glue or
picture be
If the two
them and
By contra
and part B
fusing par
The obvioeach otheis it?
The fact is
swell (dep
thus allow
with the a
note before w
is that high q
od how and w
PVC, CPVC and
– FUSING,
ufacturer of b
ence betwee
amilies of pro
y but rather th
s, (part A to p
e dive into th
tial for specif
them of the i
actors so tha
an adhesive w
elow), in esse
o components
therefore no
st, a solvent c
B. The reason
rt A into part
ous observatioer but what is
s that some o
pending on th
wing the subs
addition of so
we begin‐‐ alth
uality produc
why they wor
d ABS piping
, NOT GLU
both SC and g
n the two pro
oducts but the
he very differ
part B) which
e difference,
fiers because
mportance o
t they will alm
works by bon
ence it works
s are in actua
o bond (reme
cement does
n for this is th
B. A simplifie
on in the cases not so obvio
of the compon
he solvent – m
trates to rega
ome pressure
hough the ex
cts are readily
k and minima
systems with
ING
glues, Weld‐O
oducts. There
e difference I
rent way in w
in our case is
why is it imp
it needs to in
f not leaving
most instincti
nding itself to
as a bridge.
l contact with
mber the bon
not work we
at 80% of the
ed depiction o
e of a solventous is that the
nents (the so
more about th
ain crystallinit
pushing part
2
xplanations w
y available an
al training is p
fantastic, co
On is perhaps
e are of cours
am referring
which they acc
s pipe and fitt
ortant to und
nfluence their
the choice of
vely use the p
part A on on
h each other,
nd is only wit
ell, if at all, wh
e bond streng
of a successfu
t weld is that ere is nothing
lvents) in a SC
his later) the
ty and with it
A into part B
will be long an
d their use is
provided. This
onsistent resu
in a better po
e numerous t
g to here is no
complish the
ting.
derstand it? It
r choice of pr
f SC to the co
product corre
e side and to
there would
th the glue).
hen there is a
gth that can b
ul solvent we
the two subsg between the
C are designe
material and
t all of their p
B, now they ar
d at times co
indeed very
s is truly a ve
ults as long as
osition than m
technical diff
ot the differe
same exact t
t is in fact not
roduct or, mo
ontractor. It is
ectly.
o part B on the
be no glue le
actually a spac
be achieved w
lding looks lik
strates have nem. Did we no
ed to penetra
then evapora
physical prope
re one solid b
mplex, the m
simple, once
ry simple way
it is done rig
most to appre
ferences betw
nce in their
hing: bonding
t only import
ore important
s also essentia
e other side (
eft in betwee
ce between p
with a SC com
ke this:
now ‘flown’ inot use SC? W
te, melt and/
ate complete
erties. Howev
block!
main
it is
y to
ght!
eciate
ween
g two
tant
tly,
al for
(see
n
part A
es by
nto here
/or
ely
ver,
‘Connecti
1. Th
w
2. So
co
PART 2
It is possib
fortunate
rally work
Actually t
parts (the
functions
True Solve
We have a
what we d
These are
deeper in
always th
etc.) whe
penetratio
put toget
by figure
on the rig
to both ce
start to w
through it
Which ass
penetratio
ng’ the above
he installer ca
work them into
omehow we
ompress agai
– CHEMIC
ble to discuss
because suc
k.
he name itse
e solvents and
.
ents
already briefl
didn’t say is t
e referred to a
to the pipe a
e most desira
n appropriate
on. Assume t
her two piece
on the left be
ght. Both pene
ements meet
wiggle them–t
ts life as the m
sembly do yo
on).
e statement t
annot just dep
o the materia
must find a w
nst each othe
CAL COMPO
s the make‐up
h discussion w
lf (Solvent Ce
d the cement)
ly mentioned
hat certain so
as ‘true solve
nd fitting wal
able thing (de
e it is easy to
hat our true s
es of wood. T
elow, while a
etrate the tw
ing the ASTM
his is equivale
material ages
u think will co
to the actual a
posit the SC o
al.
way to genera
er.
OSITION O
p of SC witho
will lead to ye
ement) gives u
) which, by de
SO
that the solv
olvents are al
nts’ and beca
ll. While a ver
epending on f
illustrate the
solvent and it
The cement w
cement with
wo boards eno
M performanc
ent to the me
.
ome apart fir
3
applications o
on the surface
ate radial pres
OF SC
ut revealing a
et more unde
us the first ve
esign, will per
OLVENTS
vents will soft
so necessary
ause of their v
ry deep pene
factors such a
e difference b
ts percentage
with a greater
a smaller con
ough to initial
e standard. B
echanical load
rst? Obviously
of a SC, it is n
es (as if it wa
ssure so that
any proprieta
erstanding of
ery important
rform two ve
ten or swell th
y to actually d
very aggressiv
etration of the
as wall thickn
etween deep
e in the ceme
concentratio
ncentration o
lly hold them
But now take
ds that your p
y, the one wit
now clear that
s a glue) but
the pipe OD
ary informatio
these produc
t clue; these p
ery different y
he pipe and f
dissolve PVC o
ve nature wil
e pipe and fitt
ess, installati
p or shallow s
ent is a nail th
on of true solv
of the same so
together; th
each pair of n
piping system
th the shorte
t:
rather needs
on socket ID
on; this is
cts and how t
products have
yet essential
fitting materia
or CPVC resin
l also penetra
ting wall is no
on temperatu
solvent
hat we are usi
vents is depic
olvent is depi
is can be equ
nailed boards
m will undergo
r nail (lesser
to
they
e two
al but
.
ate
ot
ure,
ng to
cted
icted
ated
s and
o
4
Boiling Point
Some solvents are also included in the cement formulation for their high boiling point which slows down
their evaporation. This is a good feature for large diameter systems or very hot weather installations but
not so desirable for small diameter jobs where a quick cure time is needed or in very cold weather
applications.
Swelling
We have also seen how certain solvents will swell the material. This is important for two reasons:(1) it
will make the surfaces more porous, allowing deeper penetration of the solvents and (2) it will increase
the interference between pipe and fittings, which is essential to creating that radial pressure which will
in turn force material from the two components to mix with and into each other. These create that lock
and that single thicker substrate to which we previously referred.
Lubrication
In a roundabout way, solvents also act as lubricants. This is important to allow the absolutely essential
full penetration of the pipe into the fitting socket in any interference‐fit connection (a highly desirable
condition we will further analyze later). I say ‘in a roundabout way’ because the solvents don’t really
lubricate anything but create conditions the results of which is similar to lubrication. In these
circumstances, it would require incredible force to push the pipe all the way to the stop of the fitting in
an interference‐fit type joint, which would surely damage either the pipe or the fitting or both.
However, after the solvent have been worked into the surfaces, a percentage (if we did this right
approx. 1/3 of the wall thickness) of the pipe OD and socket ID will be mollified thus offering little
resistance when pushed together: they will simply ‘mix’.
Solvent Welding
In the end the right combination and proportion of solvents (accompanied by radial pressure) is what
really fuses or welds the pipe and fitting together. This is a true weld. Think about this for a second‐‐
what are the conditions necessary for what we normally think of as a weld? We need heat to melt the
parts, and we need pressure so that they flow into each other. We are performing a chemical weld,
meaning that we are achieving our melt with chemicals rather than heat, and when we add some
pressure ‐‐voila`! The fact is that if it were possible to eliminate ± tolerances in plastic pipe and fittings
we could probably get away with just Solvents and no or little need for Cement.
CEMENT
But we do have tolerances and in molded (fittings, valves) and extruded (pipe) plastic it is basically
impossible to have perfect dimensions. In addition, in North America we use tapered sockets. That
means that while we hope to always have interference at the bottom of the socket, we will always have
a gap towards the mouth of the fitting and sometimes even at the bottom. We therefore need to:
1. Fill that gap
5
2. Fill it with material that has sufficient density to act as a wedge and actually generate some
radial pressure even in places where there are gaps
3. Fill it with material which is compatible with the installation and performance of the operating
system
The ‘Cement’ components in a SC are the ones tasked with performing the three functions listed above,
and they are referred to as ‘solids’. Much like solvents, there are different materials which are used,
each with a different function.
Resin
The first obvious one is resin. This is the same resin that is used for the compounds with which pipe and
fittings are made and should have the same cell classification so that they can mechanically, chemically
and thermally perform, at the same level that the system will perform. As an example, PVC with a
pressure rated cell class should be used in cements sold for use with pressure rated PVC pipe and
fittings. With the exception of ABS (which has a much weaker chemical resistance than PVC or CPVC and
as such requires weaker solvents) the solvents used in both PVC and CPVC cements are capable of
dissolving both resins. This is an important point and means that CPVC cements can be used on PVC
systems since the solvents in a CPVC cement will soften and swell PVC and the resin deposited (CPVC) in
the interstitial spaces will be able to handle anything that a PVC system can handle (and then some). The
opposite however is not true; while the solvents in a PVC SC will soften and swell CPVC, the resin
deposited by a PVC cement will not be able to handle the operating parameters of a CPVC system.
Therefore in theory we could use CPVC SC in any CPVC or PVC application but it would be an
unnecessary expense for PVC systems.
The amount of resin that is used in a SC will vary depending on the gap to be filled and therefore the
desired viscosity, but even in Extra Heavy Body SC there is a limit (around 20%). This limit is set by
chemistry since it would not be possible (even with only true solvents) to dissolve much more resin than
that. We need to remember that this is a solution and once saturated nothing more will dissolve, just
like sugar or salt in a cup of water. It is interesting to notice that to meet ASTM a minimum of 10% of
homopolymer resin must be used; this means that two manufacturers one using 10% and one using 18%
can claim to meet the standard; but is it the same?
Additives & Fillers
But even 20% of resin will not give us the needed viscosity for SC to be used in large diameter systems,
so other materials are added. Similar to fillers, additives and stabilizers in pipe and fittings, these
materials help with various functions such as increased viscosity, improved stability and shelf life, color
differentiation, etc.
SC for che
Most of th
exception
chemical
formulate
360° inter
the only m
selected t
even a ne
not chem
leak path.
back weld
with a hot
rating, wh
specifying
The great
they are n
Pipe & Fit
We have a
weld to h
America o
The pipe O
Interferen
emical applica
hese additive
nally well in m
systems. Whe
ed with specia
rference betw
material expo
to resist the p
et fit then it is
ically resistan
. This is so co
ding of every
t air gun (nob
hich is far less
g the use of a
est SC on the
not prepared
ttings
already ment
appen as inte
our solution is
OD is slightly
nce Fit and is
ations
es are well kno
many pressure
en specifying
alty chemical
ween pipe and
osed to the ch
particular che
likely the som
nt, they will c
mmon in app
PVC or CPVC
body would u
s than any pro
proven, chem
e planet howe
correctly and
tioned severa
ended. Differe
s a tapered fit
larger than t
the ideal fit w
own and univ
e and non‐pre
SC for chemi
ly resistant so
d fitting, they
hemical flow w
micals in the
me of the sol
orrode and e
plication such
fitting. This is
se an extrude
ocess systems
mically resista
INST
ever will not d
d/or the SC is
al times the n
ent parts of t
tting socket. T
he ID of the b
which will pro
6
versally used;
essure applica
ical installatio
olid compoun
y will fuse tog
will in fact be
line). Howev
ids will be ex
rode over tim
as water trea
s a double‐sil
er to back we
s and (b) the
ant SC.
TALLATION
do a good job
not applied c
ecessity to ha
he world dea
The figure be
bottom 1/3 of
ovide the bes
; however wh
ations, they c
on, it is essen
nds. Rememb
gether around
e that of the p
ver, if there is
posed to the
me due to the
atment that s
ly move beca
eld) has no mo
real simple a
N
b if poor quali
correctly.
ave radial pre
al with this in
elow shows a
f the fitting so
st, most consi
hile they may
can be a liabil
ntial to select
ber that if the
d the entire c
pipe and fittin
s some out of
flow as well.
e chemical att
some enginee
ause (a) a bac
ore than 5psi
and successfu
ity pipe and f
essure in orde
different way
n ideal ‘fit’ fo
ocket. This is
istent results
work
ity in process
only product
re is a perfec
ircumference
ngs (obviously
roundness o
If these solid
tack and crea
ers require th
ck weld perfo
i of pressure
l solution is
fittings are us
er for the solv
ys but in Nort
or solvent wel
called
.
s or
ts
ct
e and
y
r
ds are
te a
he
rmed
ed,
vent
th
lding.
There are
Net Fit ‐ A
resistance
acceptabl
Loose Fit
encounte
With the
Fit will pro
between
will provid
fitting into
Above the
20% of th
increase o
Pipe prep
It is essen
fusion par
Although
only great
the socke
Burrs and
removed
of the pip
scrape off
e however two
A snug fit, def
e towards the
e fit which w
‐ This should
rs no resistan
proper applic
oduce joints t
pipe and fitti
de that 360° p
o the pipe.
e fusion area
e strength of
of contact are
paration
ntial to cut the
rt of the joint
an exaggerat
tly diminish t
t where 80%
filings can pr
from the out
pe or the shar
f the cement
o other possi
fined by the d
e bottom of th
ill produce ex
never be use
nce all the wa
cation of cem
that are stron
ng (aided by
pressure nece
is the bond a
f the joint is in
ea that it prov
e pipe square
t.
tion, the figur
he contact ar
of the joint s
revent prope
side and the
p edge (made
from the fusi
ble ‘fits’ that
dry pipe (befo
he socket but
xcellent resul
ed in a solven
ay to the stop
ment and prep
nger than the
our solids in a
essary to forc
area, and this
n this area, it
vides.
e since only a
re below clea
rea between p
strength is de
r contact betw
inside of the
e even worse
ion area as th
7
can be encou
ore any ceme
t which can st
ts if the ceme
t cement syst
inside the so
paration of th
e individual co
a Net Fit) at t
ce material fro
is where the
is still very im
square cut w
rly shows tha
pipe an fitting
veloped.
ween pipe an
pipe. Also a
e by the raised
he pipe is inse
untered:
nt is applied t
till be bottom
ent applicatio
tem. This occ
ocket, and it’s
he pipe both a
omponents be
the bottom of
om the pipe i
solids always
mportant bec
will provide th
at anything le
g, but it will d
nd fitting duri
slight bevel n
d edge create
erted.
to it) encoun
med out. This
on is done pro
curs when the
s actually loos
an Interferen
ecause the in
f the socket i
into the fittin
s play a big ro
ause of the v
he maximum s
ess than a squ
do so in the lo
ing assembly
needs to be p
ed by the cutt
tering some
s is still an
operly.
e dry pipe
se.
ce Fit and a N
nterference
n the fusion a
ng and from t
ole. While on
very significan
surface area
uare cut will n
ower portion
and must be
laced at the e
ting tool) will
Net
area
he
ly
nt
in the
not
of
end
8
There are actually three (3) separate reasons for not wanting the removal of cement from this area.
1. Removal of the cement will eliminate the necessary lubrication in the fusion area. This in turn
will almost certainly guarantee that the pipe is not fully inserted, thus missing the all‐important fusion
area.
2. Removal of the cement will also eliminate/shorten the exposure of the fitting wall to the
softening agent which together with the radial pressure generated by the proper interference are the
only reasons why fusion between pipe and fittings occur. A shortened exposure to the softening agents
in the cement will result in a more superficial fusion, the consequence of which will vary greatly from
decreased safety factor to shorter life expectancy.
3. If the sharp edges of the pipe remove the cement, where does it go? Only one place – down into
the waterway. Lack of a proper bevel is in fact one of the two reasons (the other being over‐ cementing
the fitting socket) why this undesirable condition can occur.
Application of cement
This document does not intend to be an instruction manual so we will glide over many of the details and
just provide a flavor of how simple steps are so important to the success of every joint.
After selecting the right cement, attention should be given to the size of the applicator. If too big, there
is too much cement is deposited in the socket; if too small, the time to coat the pipe and fitting will be
longer than the set time.In other words, the cement starts to harden before a full coat and/or insertion
is achieved.
Pipe sizes 1 ¼”and above need to always receive a second cement application on the pipe end. FIRST
APPLY CEMENT ON THE PIPE END, THEN IN THE FITTING SOCKET, AND, FINALLY, ON THE PIPE END
AGAIN. This is the brilliant ingenuity of solvent cementing!
Since we are dealing with pipe and fittings that have ± tolerances, it means that they will always be
different. We are also dealing with tapered sockets, which mean that the top, bottom and middle fit/gap
between pipe OD and fitting ID is different by design. So how do we make sure that we always use the
exact right amount of cement, not too much or too little? We simply apply a generous amount of
cement to the pipe, so that whatever gap we will have it will be filled, and a thin coating on the fitting
socket so that we achieve softening of the walls but there is simply not enough material to be pushed
into the waterway. Whatever excessive material we have deposited on the pipe will be pushed out
during pipe insertion so that (a) it is not in the waterway (b) it provides us with a very useful indicator
that all gaps have been filled and (c) we can easily wipe off. BRILLIANT!
There are many more details available (e.g. when and how to insert the pipe, set and cure times,
solvent welding in either cold or hot conditions, etc.), but again the purpose of this document is to
provide the reader with a ‘sense’ of how this simple process and products in fact require some very
specific knowledge and more attention from the decision makers and specifiers than it usually gets
9
Fusion welding is indeed a very straight forward, time proven process utilizing products with an
astounding success rate. It is also a small expense tasked with a very large responsibility. Choosing the
right products and installers who have been trained and understand how to use them is the key to
taking full advantage of the many benefits of vinyl piping systems.