infusing with contour balsa

Hi Andrew,
Before the infusion starts and once it fills the gaseous parts of the bag are run by the ideal gas law. Which is PV=nRT. N&R are constants therefore PV=T. T is nearly constant so P1V1 = P2V2. ie pressure and volume are related. If you increase the bag pressure the volume must also increase. If the volume increases then you get a poor laminate. So the aim of a good infusion is to have the bag stay the same volume or get smaller from fill start to gel. Next time I set up a bag I'm going to set up some tubes to act as manometers so I can monitior the bag pressure at different parts of the bag. Now if you have some tube at the vacuum end and when you drop volume its just a small volume of vacuum in the tube then it does not change the bag volume much, say you have a 12mm dia tube and the resin is 600mm from the job then its unlikely that the resin will retreat back into the bag and the volume change of say 100mm retreating resin is very small compared to the job size, then Its fine. (if you had a perfect nag with no outgassing then you could clamp the resin line once the resin passed the clamp and then nothing would happen, but it does so this must be outgassing or microleaks) Plus you need vacuum scavenging around the job to pick up the outgassing and microleaks so the bag stays down. I havn't been able to fully describe this mathematically but I'm working on it. There is a definite pressure profile change once the bag fills. This does not occur with bags with scavenger vac and very good bags with epoxy. But does occur with styrenated resins, probably due to outgassing at initiator sites. I haven't regularly set up bags (that I can play with) in over a year so haven't had the opportunity to investigate this further. Theres a few things on my list to check before I can accurately say "why is it so!" These effects do not happen with rigid tools, only membrane infusion. This is because with rigid tools V is always the same. As I've said before the compression in the stack is always trying to increase the bag volume so this springback is the second biggest problem after poor vacuum integrity. So when you increase your (absolute) bag pressure (decrease your gauge pressure) at the end of an infusion you must increase your volume, so this is not a good thing.

Now I said T is nearly constant but when the job exotherms it clearly is not so when the job heats up any gasses must heat up and they must increase in volume, so any bubbles you can't see at room temp you will see at 40-50-60deg if it exotherms before gelling. This maybe the simplest answer to why we get visible bubbles simply because of the resin heating up... Anyone with more ideas?

We don't see these bubbles in autoclave parts because we are using a positive pressure of say 6 bar which keeps the bubbles compressed until gel. In rigid mould VARTM we have a trapped laminate, a fixed volume so once the resin flows through you can disconnect the vac line and resin line and open it to air pressure and walk away. Its just like filling up a cup of water at that point. Can't do that with membrane infusion because of the stack springback.

The difficulty with standard infusion is that we use vacuum to compress the stack, to provide a pressure difference to transport the resin, to remove air from the stack to allow the resin to fill it easily, to create a low pressure to remove water vapour to name the important ones. So we use the same vac for many things some of which are in conflict. This is where double nags and rigid moulds come in which are trying to seperate the functions of the process to allow better control. Its a pure pysical process so once we understand thge variables and the process it should work the same every time, every time... every time.. The current project I'm working on they built an air conditrioned room so it runs at 25degs all the time. It has improved the laminates visibly and made the process very predictable. So temp is a big process variable.

Cheers Peter

 

The best reason I've read 2 that advocates the reduction in vacuum after infusion is this;

When the resin line is opened, you have atmospheric pressure equal inside and outside the bag. Wrong, you have atm outside, also atm inside PLUS 150mbar laminate spring pressure. This then allows A pressure imbalance and the stack must increase volume. Once the part is infused, the increasing volume allows any trapped air to expand and the voids will grow in size. After clamping the Inlet if the vacuum is reduced by 150mbar then the imbalance is equalized and the expanding air will not push resin out the outlet and the bubbles will remain the same size.

There's a lot to digest in the above so I suggest reading it several times as I've tried to be very concise writing from my Android. ...

 

Lets break Gropers advocates process down, I think its totally wrong.

The advocate needs to learn about gauge pressure vs absolute pressure to begin with. eg if at resin fill start there is 1150mbar in the bag it will be blown up like a balloon not compressed. I'm trying to expore this not just debunk it... The other logic does not make sense either so you will have to explain further please. eg once the resin flow is stopped the bag increases volume. Why should it increase volume? my bags don't increase in volume when I clamp off the resin? to increase in volume, as the resin can't gain volume so it has to be inflow gases from leaks or outgassing from resin.

This is how it should go. Do a std bag and start the fill. At part finish the resin "in" tube is clamped or valved, then the vacuum line now containing resin is clamped or valved off. Now we have a given volume of resin in the bag. If the resin didn't outgas or there were no leaks this would be the finish. It would sit there until the resin gelled. Even the springback couldn't increase the bag volume as its hydraulically locked, resin is incompressible and inextensionable. So the only thing can be outgassing or leaks?? that increase the bag volume, hastened by the stack springback.

more thoughts ???Peter S

 

Just thinking out loud, Bag "springback" Could it be reasonably quantified using a dial indicator initially on the contact surface to an overlying plate & then to the outside of the bag using the same plate to give a calculatable volume difference? Also , & I only heard about & not witnessed this, a pre-eminent Oz builder doing some other "simple" & not so "critical?" jobs such as or similar to large door components was "charging" the bag envelope with the resin source from/at height giving a head of maybe 20+ feet.... not sure what this achieved or was just to expidite or solve symptoms of thicker resins?
Jeff.

 

Hi Waikikin -
A) there are many studies done that measure the change in thickness during an infusion. If you do some searching you will find them. The crux is that in a good infusion the bag gets thinner as the resin flows by due to the resin acting as a lubricant letting the external pressure compress it a little more.
B) if you put the resin as little as 200mm above the bag it will syphon and the local area around the resin inlet will expand. Syponing flow rate is generally faster then infusion fill rate near end of bag fill and is very undesirable. 20ft would generate a hugh static pressure and just fill up the bag in a resin mess C) springback does not occur if the bag vacuum integrity is good and there is adequate vacuum scavenging in the system ie there is a vacuum circuit around the job maintaining atmospheric pressure on the stack. Cheers Peter S

waikikin - heres a thesis on laminate compression. Doesn't give any help for infusion in general but does investigate the visco-elastic behavoir of the stack.

 

Things to know -
A) Ideal gas law runs the gas volume side of the process, assuming the volumes are static. So need to look up the difference between a static vacuum space and a dynamic vacuum space. Look at your vac pump performance curve to understand the difference
B) darcys Law runs the resin filling process, this is accurate and totally predictable
C) Manufactured leaks - are due to poor workpersonship and poor consumables
D) micro leaks - are due to porous consumables, valves, etc and various tacky tapes. Good technique reduces these to a very low value, a value which is just perceptable over a couple of days.
E) outgassing is a function of surface tension, bouyancy, critical bubble size and local pressure. We can have micro bubbles and saturated gases. There is a thing called the critical bubble size, if the bubble is less then the critical size it is trapped and static. If it grows to critical size it expands and then rises due to bouyancy and goes to the surface (with surface media infusion these over critical bubbles get to the media and stay there.) The sub critical bubbles stay in the laminate. I have played with surface tension modifiers and defoamers and they do not help with infusion. This is because these additives are bi-polar molecules. One side lives in the resin and one side sticks out of the surface modifying the surface shape hence tension. Unfortunately the bit sticking above the surface generally interacts with air moisture so in vacuum this is no good. I have overdosed resins with very expensive modifiers and cannot see any difference
F) Degassing works, but this is doable with small quantities not very large things like boats. It is possible to heat resin and degas it. Epoxy responds very well to this as it does not have any solvents. So heat it to 50-60deg and under a small vacuum it degasses really well. If you heat styrenated resins, the styrene boils off really fast. But styrenated resins are low enough viscosities that they degass really easy anyway.
G) Initiators are everywhere in an infusion stack. So its inevitable to get the champagne effect happening ie spontaneous outgassing
H) Its imperative in a standard infusion to match the fill time and the gel time. The longer the wet time the more chance we have of the outgassing and microleaks to degrade the stack condition. If we had a cure on demand resin we could use simple VIP process and as soon as the resin fills push the cure button and it would be perfect. This can be done with various techniques, heat, radiation, UV
I) Its apparent to me that around 95% vacuum we may not need to degas and we may build visually clear laminates. The problem is maintaining a stable dynamic vacuum vs a static vacuum or near static vacuum.
J) There seems to be a volume ratio effect in the process. eg its really easy to make small scale exceptionally high quality laminates. Use the same process but scale up to a boat/car/plane and it does not work as well. It seems to me their is a vacuum volume to stack volume ratio which is important. This can most easily be explained in say a small job 1mx1m using a 200mm flange produces great laminates all the time. But to do the same with a 20m boat part and it doesn't work as well
K) Water vapour - when water turns to steam it gains volume by 10,000x. It takes a while to steam off water from jobs before you start to fill . If you fill before the water is removed this contributes to voids. I really didn't get an appreciation for this until I used an absolute vac guage that can run continuously. You see the pressure drop then stabilise say at 500Pa and steam comes out of the vac pump. About 45mins later the steam stops and the gauge starts dropping again. Dry laminates make stronger laminates
L) Temperature is a huge variable and can change viscosity and gel time simultaneously. Big problem. If possible use temp controlled room or tools. water controlled tools are best as you can heat and cool them. Look up light rtm tool design with copper tube cooling/heating
M) Ultimately any vacuum process will produce porosity. So need to progress to positive pressure processes. These involve double bags, rigid tools or combinations. There is much work in this area now
N) What else can I think of? semi permiable membranes make great resin brakes and vac port isolators.

Summary: pay attention to vacuum, dry the laminate, remove all manufactured & micro leaks, degas if possible, match gel time to fil time, get better at everything as time goes by.

Cheers Peter s

 

Back to what I said earlier. I used to believe what you essentially said. Then I was given an explanation by the maker of MTI hose in Germany.

In a flexible bag setup, with vacuum applied right the way through to cure, the reduction in vacuum by about 150mbar can reduce the growth of voids after clamping. In the real world , achieving 100% vacuum integrity is very difficult, so isolation of vacuum after clamping isn't a practical method for a lot of us doing out of laboratory infusions- you know what I'm getting at.

So with the vacuum acting on there laminate until cure the following applies.

The bag does spring back a little, if it cannot, then the pressure differential is still there. the measured force of a typical laminate spring pressure is measured at about 150mbar, many references agree on this. if it didn't, you would not get a resin rich area near the inlet or a difference in the resin content across the laminate between inlet and outlet. This evens out slowly after clamping as the viscosity resistance prevents it happening instantly.

If the vacuum is maintained at 100%, the pressure imbalance of 150mbar will be still present and will effect anything that is elastic or compressible etc. thus the air bubbles grow a little and to maintain bag volume, some of the resin is pushed out the vac line by any expanding air trapped in the bag. If there is no trapped air, then there is no problem, but some air in solution is always introduced via the resin in non laboratory infusions where it cannot be degassed 100%. This is where the MTI hose or semi permeables come in, they mechanically isolate the means for the resin to move out of the bag and thus any bubbles present cannot grow as they are hydraulically locked as you put it.

Took me a while to get my head around the whole thing, but after a while it began to make sense.

 

Hi Petereng,

Thanks for the info & link,
On the siphoning flow, rate is this dependent on the siphon flow being unregulated/"throttled" back particularily towards the end of wetout. I'm quite happy to accept that having the resin bucket lower is good-seems to be regular practice, the example I was given may well have been a red herring to protect actual practice although the rep that quoted it was there & the job largish... maybe they shot the juice in quick, clamped off & let it spread? was 10 plus years ago.
So if the rate of resin volume flowing through the stack isn't exceeded by the supply flow & vac level exceeds supply pressure from resin feed the bag stays down & tight? From your link .....
" The most important manifestation of this viscoelastic
behaviour is the significant stress relaxation which occurs, observed by many investigators, e.g.
[16-17], and thought to be due to frictional fibre slippage. The stress relaxation response implies
that the stress required to hold a fibrous material at a constant volume fraction decreases over
time, and this can have consequences for the tooling forces in the RTM and I/CM processes,
particularly since the observed stress relaxation often occurs over time scales on the same order
as that of a full LCM process. Similarly, the observed thickness fluctuations during the flexible
bag processes are due in part to the viscoelastic nature of fibrous materials [18], and an accurate
simulation of these processes requires a time-dependent model of fibre compaction. A number of
authors have proposed that an elastic model be employed for the compaction phase and a separate
model to capture the relaxation behaviour once compaction has ceased/paused"

& the stack relaxes & squeezes down some more. The big variable is the membranes relationship/thickness distance to the mold/contact surface?

I'm just doing quite a simple plug/then mold to do a roll edged bench top(2420 x 620 x 52) & thinking it's probably a nice simple thing to practice some infusion on, I had planned to practice on some simple molded engine boxes but did a quick melamine/plastcine mold that wasn't vac proof & just laid them up. These bench pieces form topside stinger units & include some uni tape but I may have to tab these later as they're drawn at 90 degrees to bench top & will have to fill the curl on the J part of the bench front with foam.... may be able to include in infusion though, the rest is solid glass with gelcoat finish.
Jeff.

 

Peter; thanks for taking the time to pen the detailed information. While I dids not say so in post #65 I too was using the ideal gas law to support the practice of reducing vaccum at completion of impregnation. What I did not say and should have is that it requires stable vacuum. I took this for granted as good infusion requires a stable vacuum at what ever your targeted vacuum is, be it 80% or 100%. This is why I argue against the practice of having automated switching on and off of the vacuum pump as proposed in previous infusion threads.
I think that we all agree that after a bit of practice and attention to detail you produce laminates that are far better than wet bagging. But to do just that bit better to minimize micro voids I considered:-
1. double bagging.
2. use of additives such as antifoams and surface tension modifiers.
3. use inline static mixers for the resin. 1,2&3 have not done to date as I do not see the cost to benefit for me to be worth it.
4. deggasing; I do this for smaller parts as you pointed out earlier it is easy to build a cheap vacuum chamber that will hold a 10L plastic bucket. Yes use of temperature helps speed this up.
5. reduce vacuum after impregnation. For me this is simple and easy to do as I mentioned I have a small diaphragm pump with ~85% vacuum capacity. The fact that I am running this pump at its full capacity rather than reducing and having to control a high capacity pump is why I think I have a stable vacuum that is required.
The explanation why I think this works is the ideal gas law acting on micro and sub critical bubbles.

Cheers
Andrew

 

Hi Groper - Where is this discussed on the MTI site? I can't find it. They do mention that <0.12psia is the optimal pressure for VIP several times. 14.5psia is 1 bar so 0.12/14.5x101000=836Pa (or 8.36mbar or 99.2% vacuum) which is a typical infusion process pressure. The springback pressure of a laminate is considerably greater than 150mbar otherwise we would only need 150mbar relative (0.15atm or 1.5 T/m2) to compress the laminate. The MTI approach still has problems in that the bubbles still have to "transport" to the tube to get out. At fill end when you close off the resin and the resin stops flowing, the bubbles cannot be transported to the MTI from say 100mm away, so only the local bubbles very close to the MTI product get sucked away. Its a great product if you have small flanges and can't have space to set up breaks.

The resin richness at the resin feed in is as you say a function of the viscosity/friction and time. If the consumption rate of the laminate and the feed in rate are the same this does not happen however. If there is considerable resin flooding at the resin feed in this points to poor vacuum scavenging of the job, not a "natural" part of the process. You can acheive the same result of the MTI tube by placing spiral wrap around the job with a space of at least 50-75mm away from it, with the bag coming in contact with the tool or have glass tissue or even newspaper in the gap. Bubbles will travel across this gap and get scavenged by the spiral (I use cheap 4mm rope, cheaper then spiral) just as well as the MTI tube. (mdf board is good too its slow to fill and maintains vacuum communication until the resin gels)

The volume change of the job from outgassing is so small it does not come into the equation ie it can't possibly push resin out of the job as its only a very very small volume.

There is a difference in two processes I've noticed. 1) You close off the resin feed then close of the resin outlet. When the resin flow stops outgassing appears, if you start flow again the bubbles disappear 2) If you use a small resin outlet hose say 4mm and run this to a resin trap and let the resin flow through until gelled then porosity does not appear 3) lite rtm uses this approach. RTM lite only uses 50-60% vacuum pressure so does not suffer from water vapour, outgassing or other problems. Because the stack is lightly compressed they fill fast. The outflow is into a resin trap and it flows until gelled or some people clamp off. Since the mould is rigid it is hydraulically locked so outgassing does not occur (in conjunction with the "high" process vacuum) Plus it is a perimeter feed approach which has high merit and is being used more now in std VIP.

If you are increasing pressure (please discuss in absoulte terms not relative terms) at end of fill and you have resin in a tube then I'm not concerned because this is only a very small change in the system. If however you are increasing pressure in the bag then this is a serious physical problem from a physics point of view. Hope this helps.

A VIP system is no different from sucking water along a pipeline with some part of the pipe being clogged up and some parts of the pipe are soft plastic so can change shape. There is a pressure profile along the pipe due to hydrostatic, dynamic, voscosity and friction effects. It seems that things go quite well when there is flow and when the flow is stopped is when we get into trouble.

addition- outgassing in styrenated resins can be from the catalyst. One part of the reaction is the evolution of hydrogen and this could be a considerable amount. Are there any chemists in the ranks to comment please? CHM type catalysts are best as they claim no outgassing

Hope this babble helps. Peter

 

Hi Andrew - If you are infusing at 85% vacuum (15000Pa or 150mbar absolute) then you have many advantages I think. You have higher porosity so the flow length is very good, you are way above the outgassing and water vapour problem, so the only problem you have as I see it is that if you infuse too fast you include the air at the flow front quite easily (beachfront effect V turbulent flow). So a slow fill with good vac scavenging will produce excellent results. I'm aiming at high structural performance laminates which require high vacuum to compress the laminate to acheive 30-35% resin by weight with fibre fractions by volume around 55% glass and 60% carbon. At 85% vac its unlikely to acheive this. ie equal to or better then a very good prepreg. Can you tell me the thickness and weight of some of your laminates so I can calculate their ratios? I need to know the resin and the laminate weight and the thickness. Don't worry about surface tension modifiers I've spent a lot of time on these and can't see the difference. Spent a lot of time with the Dow chemists, used 6 different types etc etc. On big jobs like 50ft cats (Perry cats) we used chopper gun pumps and they worked a treat, only ever had <10 litres of mixed resin at any moment. When you have to deal with a shot of 300-400kgs of resin, best to mix on the go. I was involved with the first 45 footer to be infused at Riviera and they had nearly 200 litres of mixed resin and to me it was really scary if something went wrong as they had no backup plan to deal with that situation!! Cheers Peter S

 

Dont get confused with the MTI process here, i only mentioned the name because this is where the explanation came from.

We are not talking about bubbles getting transported away, only bubbles that remain in the laminate, well away from the resin front, growing in size. This growth occurs when the resin line is clamped off and the vacuum outlet is still open, waiting for the resin to cure. Increasing the absolute pressure at this time, can reduce the bubble growth by the direct mechanism of equalizing the pressure difference - which occured in the first place by the spring pressure of the laminate + the atm of the open resin inlet. The total pressure inside the bag during resin feed, in the fully wet area some ways behind the resin front, is greater than the 1atm outside the bag.

This is the only point i was trying to convey...

 

Nope... infusing at less than 100% is never an advantage, and andrew K is not advocating this... Its the reduction of vacuum pressure (increasing absolute) AFTER infusion, that he is talking about - SAME AS IM TALKING ABOUT!!!

Its like the engineer and manager flying a hot air balloon story - heard of it?

In case you havnt, here it is;

A man in a hot air balloon realised he was lost. He reduced altitude and spotted a woman below. He descended a bit more and shouted,
"Excuse me, can you help me? I promised a friend I would meet him an hour ago, but I don't know where I am."
The woman below replied,
"You're in a hot air balloon hovering approximately 30 feet above the ground. You're between 40 and 41 degrees north latitude and between 59 and 60 degrees west longitude."
"You must be an engineer," said the balloonist.
"I am," replied the woman, "How did you know?"
"Well," answered the balloonist, "everything you told me is technically correct, but I've no idea what to make of your information, and the fact is I'm still lost. Frankly, you've not been much help at all. If anything, you've delayed my trip."
The woman below responded, "You must be in Management."
"I am," replied the balloonist, "but how did you know?"
"Well," said the woman, "you don't know where you are or where you're going. You have risen to where you are due to a large quantity of hot air. You made a promise, which you've no idea how to keep, and you expect people beneath you to solve your problems. The fact is you are in exactly the same position you were in before we met, but now, somehow, it's my fault."

 

Groper and Andrew - Do you have a column of resin in the vacuum tube when you increase vacuum line pressure? If so I'm happy about that because the consequence is very small. If you have an arrangment in which you are controlling the bag pressure (vs the resin exit line pressure) then you have a problem because the bag must increase in volume considerably. So can you show me a picture or drawing of how you are circuiting the job and where you are clamping off and controlling the line pressure? Forums and discussions like this are difficult just like the balloonist story. Just need to persist until we figure it out. Peter

Groper there is no vacuum pump commonly used for VIP that can produce 100% vacuum. 99.9% (100Pa double stage pumps) yes, 99% (1000Pa is common) most boat builders I've measured are at <98% (>2000Pa). I just say this because if you have a bag at 98% then there is a huge amount of air in there that can get trapped during the infusion. Don't need to consider outgassing, leaks or anything else its right there in plain sight (but invisible). Cheers Peter enjoying the discussion has given me some stuff to investigate further thanks!

 

In short, yes, the guage pressure is reduced to ~850mbar after the job is filled. As i referred to earlier, this equalizes the pressure imbalance from the laminate spring pressure and any remaining air entrapped in the laminate will remain the same size - hopefully very small, almost invisible etc rather than increasing bubble sizes if the 1000mbar is held until cure.

Some more details;

Peel ply resin brake acts the same as a semi permeable fabric - allows air through but resists resin flow. This disables the mechanism for bubble expansion after infusion in the same way as MTI etc.

I use a 2 stage HVAC vac pump - as i work in the refrigeration industry... My pump is rated at 25 micron absolute. I never get this far because i always have some form of leaks somewhere, very few times i get a perfect seal, which seems to be easier to achieve on smaller jobs as opposed to big jobs. Infusing small jobs is a waste of time tho, i prefer to infuse the biggest laminates i can as its more time and labour efficient. The same pump routinely pulls below 200micron on refrigeration systems prior to charging with the refrigerant, as we always MUST get a perfect seal in this instance.

I always notice a stall in the vacuum pressure as the moisture boils off, then it continues to pull down again until the leak integrity prevents it going lower. This usually takes an hour or so for my conditions up here in the tropics with a 4 hour setup time with the preform open to the atmosphere.