(Prime 20 LV infusion Epoxy) foaming under vacuum

[otseg]

We have been unable to solve the problem of the epoxy foaming under vacuum when it hits the feed lines under the bag. The foamy epoxy goes into the top laminate stack and stays there. We get a 6' to 16" wide white streak under the feed lines and the laminate compressive properties are compromised. We degas the first 10 gallons of resin. After the feed lines fill we don't de gas all of the resin, as we have found that once the feed lines fill its not a problem.

Anybody have a silver bullet to this problem? I'm ready for a lead one for my head.

[KnottyBuoyz]

I was going to suggest degassing the resin before you start the infusion but you got that. How long do you keep a full vac on the layup before you start the resin flowing? Are you using a flow media? I've had similar results on much smaller parts. I let the resin lines fill to the point where it's just about to enter the stack and stop it for a few mins. This'll bring the layup back down to full vac before I let the resin loose into the layup. I've used heavier (Aerotech red) flow media that seems to trap a lot of the foaming resin and keep it there.

[otseg]

I was going to suggest degassing the resin before you start the infusion but you got that. How long do you keep a full vac on the layup before you start the resin flowing?

2 hrs for a small part, over night for a large one.

Are you using a flow media?

We are now, it helps soften but not eliminate the problem. it simply spreads the foamy resin out over a larger area.

I've had similar results on much smaller parts. I let the resin lines fill to the point where it's just about to enter the stack and stop it for a few mins. This'll bring the layup back down to full vac before I let the resin loose into the layup.

Thats excellent. We found out the same through trial and error. We call it burping the lines.

I've used heavier (Aerotech red) flow media that seems to trap a lot of the foaming resin and keep it there.

We tried both, and as far as minimizing the foam, you still have the foam. We use the green because it flows faster. This is a shot of infusion that utilizes the grooved core, no flow media. If it had been with flow media, then the white would have been softer/wider.

[KnottyBuoyz]

Is that a wood core or foam?

Most wood is porous and will leak. Try enclosing the entire part in a bag for a perfect vacuum.

Once you "burp" the resin lines try releasing it slowly to start and once it gets moving then open them up full.

Are your feed lines spiral wrap? If so try putting a piece of peel ply under them and on top of your layup. The peel ply will let the resin through to the glass and maybe stop the foamy stuff from hitting your layup. Just a though.

Have you got any pictures of your layup before you start the infusion? Maybe re-routing the vacuum & feed lines would help.

How exactly are you de-gassing your resin?

[otseg]

Is that a wood core or foam?

Its H-80 Divynicell

Most wood is porous and will leak. Try enclosing the entire part in a bag for a perfect vacuum.

We use a precision Vacuum guage and our drop tests are close to perfect.

Once you "burp" the resin lines try releasing it slowly to start and once it gets moving then open them up full.

Substantially we do, however a small part is 10' long, a large part might be 50'

Are your feed lines spiral wrap? If so try putting a piece of peel ply under them and on top of your layup. The peel ply will let the resin through to the glass and maybe stop the foamy stuff from hitting your layup. Just a though.

They were Omega flow extrusions, then we switched to spiral wrap and now we have a product we make ourselves. We have tried all types of laminate stacks under the feed line. Works like a couple layers of Kevlar would stop a 9 mm. We have also laid the feed line on a 100 oz of fiberglass laid on a perf. Helps a little.

Have you got any pictures of your layup before you start the infusion? Maybe re-routing the vacuum & feed lines would help.

see below

How exactly are you de-gassing your resin?

4 gal at a time in a big chamber full vacuum for 10 minutes. The resin and hardner are dispensed through a static mixer. We are not introducing air through mixing.

[Stumble]

Otseg,

It sounds to me like you are experencing the effect of being to good at your quality control. :P

I know it sounds weird, but by mixing the resin in a way that prevents bubbles from being introduced you effectively eliminate the ability of the resin batch to actually be degassed. The problem is that there are two types of gass molocules in a batch of resin, the first is the small bubbles usually caused by mixing air into the batch these are easily removed by exposing the batch to a vacume and it sounds like you are doing this. The problem is the second state of molecules in a batch.

This is the gass molocules that are actually disolved into the liquid, and these will not be removed by simple exposure to a vacume. It requires a nucleation site for these molocules to be pulled out of solution not just a vacume. In a simple vacume degassing process the mixed in bubbles act as nucleation sites and help pull the disolved molocules out, but you are being careful to NOT create bubbles preventing this from occuring. So your quality control here is actually hurting the finshed product since the lack of bubbles reduces the ability of the disolved molocules to be evacuated. This of course leaves you with a supersaturated batch of resin with molocules of gasses just begging to be released, add in a little turbulence from being forced through the infusion process and the resin foams.

The trick then is to try and get this supersaturated resin to give up the molocules it is storing, and that can be done in a number of ways. Either purposefully indice lots of small bubbles to act as a nucleation sight, through a process called sparging, or add a filter that is good at causing nucleation to pull the molocules out of solution.


For a technical paper on exacally this problem see http://www.lightweight-structures.com/index.php?option=com_content&task=view&id=34

[otseg]

The scotch pad as a bubble initiator I heard about earlier today for the first time. Initially I built a vacuum chamber to hold a bucket with the chamber top that had a gland into which the static mixing nozzle supplying the resin fit. Resin pissed through the vacuum on its way to fill the bucket. It was probably effective, however in another test I also watched a full bucket of mixed resin bubble so vigorously on its own, I concluded that the dispensing chamber was unnecessary.

I could easily insert a bubble maker with air injection as a component to the static mixer, but the supplier was adamant that the resin should not be aerated.

Yes we are all about QA procedure as we achieved our DNV approval, are working on our ABS composite certification which I formally had for my previous facility and are seeking to achieve an ISO 9002 registration at year's end.

Functionally we solved the problem by building sacrificial laminate discrete from the real part. That got us though the parts we needed to make. Sometimes I would get cocky thinking I had the problem licked with some new procedure and then get a totally unexpected comeuppance. I need to study the link you provided in detail. I really had hoped that this was not my problem to solve.

My guts tell me that solvent based diluents are used to reduce the epoxies viscosity for infusion, but the supplier swears it's not. I was prepared to build 1000 Kg capacity resin and hardener vacuum storage containers to pre degas the components, but was warned that needed components would be bled off.

[AndrewK]

Otseg,

I have experienced the same but not to the extent of degrading the laminate as shown in your picture as it is being trapped in the shade cloth used as the flow media.
In my case I think that it is the vapor pressure of the hardener that is the problem. The reason why I say this is that the problem is much worse if I blend fast hardener with the slow. With the slow I only notice it if the temperature is >35'C.

I placed small clear glass jars containing the fast, slow hardener and resin into a large clear jar and pulled a vacuum on this.
After the initial release of dissolved gas the slow hardener and the resin had a gas bubble form say every 30 sec, the fast hardener every couple of seconds, clear difference.
Vapor pressure is temperature and pressure related, if you can not resole it by other means you could try reducing the vacuum say 90% after the infusion is complete.

[otseg]

I was keeping my molds and resin at a constant 82 F, winter and summer. A fellow I met found that he could endlessly make the resin outgas again after degassing by raising the temperature in 10 degree intervals. Higher temp, more gassing. Maybe it would help to keep the resin cooler.

I did a part and infused it at 600 millibars. It seemed to not gas. then I cranked the vacuum to 20 millibars and like magic the dread white streak revealed itself. So what's going on with that??

[JRL]

After looking at your custom made resin sock picture I think I might have your answer.

Your problem may be as simple as NOT doubling up on ANY of your flow medium that comes in contact with part. When you do this you create areas of low pressure that will attract any vapor/gas left in the resin. Since the areas to the right and left of the sock have more "pressure" your gas will migrate to the sock because of its lower pressure (essentially a bridge because the bag can not pull through double stacked FM). If you look closely at your pics there are signs of trapped air in the FM sock.

Heres an example of my old nightmare. The part on the left was infused at 27.5 hg. All FM was butted instead of over lapped. The part on the right was layed up identically, only difference was over lapping FM. Both bags had absolutely ZERO leakage. This was a relatively thin layp.... 9 layers of 8.8oz glass, and max thickness of 13 layers at the base.

http://www.photohost.org/gallery/data/500/107Resin-Infusion.jpg

[otseg]

The green thing you are seeing was used as a pre feeder to the primary omega flow resin feed. It was a spiral wrap, wrapped in flow media. We were trying to let the gas diffuse before it got to the primary feed line, fill the primary feed line and then have both open. It was one of a hundred things we tried. Didn't work.

When the resin hits that big empty feed line, it foams. once it foams, there is no getting the foam out. We use the moving resin front to fill the next feed line before opening it, and that works ok for the follow on feed lines.

What is happening to your part? It looks like the feed line was closed off and the part got sucked dry.

[Stumble]

Otseg,

How tied to this resin supplier are you? It sounds like their tech support people can't give you any ideas of how to solve this problem, which is usually the point where I start looking for other suppliers.

[Stumble]

It looks like lead is going for around 70 cents a pound on the world market so you will probably get about half that for a keel at a scrap yard. For this much in a mass it would probably be worth calling around to a couple of yards to get prices though.

Add in the ebay value of the winches and deck hardware, and if you are willing to take a chainsaw to the boat it might be able to help fun a more reasonable project boat.

[AndrewK]

Otseg,

If the problem is only in the first distribution line then at the end of the infusion turn this into a vacuum line and suck all of the gas out.

I do not think your issue is dissolved gas but something in the resin mix boiling at the low pressure, same as what I get with the fast hardener.
Check all of your components as I described, could also be due to wet resin and water boiling off.

[otseg]

If the problem is only in the first distribution line then at the end of the infusion turn this into a vacuum line and suck all of the gas out.

Thats effective when someone kicks a hose loose or the feed line runs out of resin. The part gets a big gulp of air, and if there is a flow media it tends to stay on top. You can suck it back out.

This foam is the first resin to hit the part and it has shown no inclination to migrate once it is in the laminate. Prove this as you shoot a part switch to a tinted resin and you will see that the follow on resin does not push the first resin along the part, but takes the path of least resistance and wets out the dry stack, leap froggin over the first resin.

I do not think your issue is dissolved gas but something in the resin mix boiling at the low pressure, same as what I get with the fast hardener.
Check all of your components as I described, could also be due to wet resin and water boiling off.

I think this is fair to say, however we take great pains to keep the materials dry and the part always has a bag over it if not immediately being worked on.

The question may be are we over doing the vacuum achieved? Is there test data for laminates shot at 200, 400 or 600 millibars instead of 20-30 millibars we achieve?

[AndrewK]

Otseg,
If you do not want to change resin then you can only play with temperature and pressure to control the boiling. Since you control the resin and mold temperature looks like you need to establish what the critical pressure is and infuse just below this threshold.
I am not clear on what is happening as you say that at constant 20 millibar the gassing is only occurring at the first feed but not at the subsequent ones.
Then you also say that when you infused at 600 millibar (400 millibar vacuum?) there is no gassing but as you increase to full vacuum gassing begins as expected.

[otseg]

We use the first feed line to fill the the next adjacent feed line completely and actually a foot beyond. The diference then is that the first feed line is empty under the bag, the subsequent feed lines are full. This helps greatly.

If the feed lines after the first ones are not pre filled, then they foam just like the first one.

I shot a boat a week previously in VE and had not this issue. I can't say I am an epoxy neophyte as I have been building various boats with epoxy since 1969, switched to VE, and now back to epoxy, principally for the properties, one of which is that the very high flash point of epoxy settled down a very twitchy fire marshal

I never wanted to be the lab rat, but it seems like I need to make a series of tests varying the limits of vacuum and temperture that this particular product can be shot at. I invested a $1200 in two precision vacuum regulators to put in my system. I don't think just bleeding off the catch pot would be control enough.

As I said before, starting with less vacuum at the initial fill and then pulling it down did not work.

[AndrewK]

"As I said before, starting with less vacuum at the initial fill and then pulling it down did not work."
This is what I do not understand, if the subsequent lines get filled without gassing up why would the first line once filled at lower vacuum also not remain filled ie gas free when at the same vacuum.
The only thing I can think off is that the pressure in your subsequent lines is higher than the initial 20millibar. Have you tried connecting gauges to all lines to see what the pressure distribution is during the infusion?

[otseg]

I have no idea why the first line when filled with the part at 400 to 600 millibar's would foam when the vacuum was increased. I know I don't do that any more!

The vacuum on the rest of the part is only less from the first moment when you open the feed lines.

I shot 5 parts today at 300 millibars. These were small parts, ie console, door and leaning post for a power cat we are putting in production. They appeared to infuse normally. I will see when we strip the bag. I wish there was some test data to be shared as to laminate properties shot at various vacuum levels. I did not vary the vacuum level on these parts as they gelled. (see lesson learned # 759 above) Its under vacuum over night. We will be doing the hulls later in the week.

[AndrewK]

Otseg,

"The vacuum on the rest of the part is only less from the first moment when you open the feed lines".
If the second line is not foaming I was suggesting that the vacuum was much less then the initial 20mb abs pressure, although I would not have expected it to be. It would be good to know.
Before I discovered that the fast hardener was my problem I experimented infusing at 400 & 200 mb abs pressure, I found the rate to be noticeably slower.
I have not seen any comparative laminate data for infusion at various pressure.
If resin content is approx 50%wt with no vacuum consolidation and 70%wt with 20mb then the resin content at 200mb will be somewhere around 65%wt. So the laminate properties would not be too different from what you have been getting.

You mention that you make your own distribution line, are you willing to describe this.
I use spiral wrap and would like to avoid the imprint it leaves if I could.
I had thought about making my own version of omega lines out of molding silicone but have not tried yet.
Is it worth attempting?

Andrew

[otseg]

Supplimental feed lines are quite happy to foam if not sustantially filled by the resin front.

Wrap the feed lines in flow media and lift the body of the line off the laminate stack as you are pulling down the laminate. We did this as a method to possibly difuse the foam before it hit the laminate stack. It helps, plus it keeps the exotherm in the line off the part

I went through a thousand feet of silicone omega and we found the vynal extrusion version to work as well. We were diligent in applying release to it after each use, it stripped fast but I had one guy who did nothing but clean and apply release to it. I don't use it any more.

[Richard Atkin]

Who said boatbuilders are always looking for ways to rip off the customer? Nice to see all you guys care so much about quality.

[otseg]

I shot 5 parts today at 300 millibars. These were small parts, ie console, door and leaning post for a power cat we are putting in production. They appeared to infuse normally. I will see when we strip the bag. I wish there was some test data to be shared as to laminate properties shot at various vacuum levels. I did not vary the vacuum level on these parts as they gelled. (see lesson learned # 759 above) Its under vacuum over night. We will be doing the hulls later in the week.

We did not experience any problems with these small parts. Nor did we with the hull. One variable is that the laminate was substantially single skin. Another is that the other parts we made were all cored and laid up on a temperture controlled (85 F) flat mold. Since the molds for the small parts and hull were ambient temp, perhaps the higher temp and higher vacuum for the other parts are more condusive to boiling of the epoxy components.

http://www.forumpictureprocessor.com/pictureprocessor/images/IMG_1190_1.jpg

http://www.forumpictureprocessor.com/pictureprocessor/images/IMG_1209.jpg

http://www.forumpictureprocessor.com/pictureprocessor/images/IMG_1550_3.jpg

[wigentd]

I saw your thread on boat design
http://www.boatdesign.net/forums/materials/prime-20-lv-infusion-epoxy-foaming-under-vacuum-28395-2.html

Did you ever solve the foaming?

We have seen similar results with sc-15 epoxy and carbon fibers recently.
Our R&D staff ran some vacuum trials a few weeks ago indicating nucleation of gas from solution at room temp and -22-25inHg vac or so, but only in contact with fiber.

I was particularly interested in your description of the difference between the primary inlet strip filled from supply, and the follow on strips filled through fiber from primary.
Also, the invisible foam scar that does not bloom under the flow strip until the vacuum is raised is interesting.
Together, these two phenomena may hint at mechanism.

While tumble/mechanical mixing seems to the main difference, there are more differences between resin in the 1st strip and the 2nd 3rd ….

1st strip 2nd 3rd …

System vacuum resin supply vac
in dry pack and at flow front vacuum pump dominates because gasses migrate to exhaust much faster than flow front advances
,,,,just behind flow front, resin side friction (or lack of) determines if vac state is closer bucket head or vac system setting

fast fill strip slow fill strip
poured over top fast wicked edge-in slow
virgin pocket resin degassed through fiber

dumping virgin resin (never touched extensive nucleation surfaces like fiber) quickly top down through thickness both traps air between filaments under strip, and also leaves a pocket of virgin resin under the 1st strip…
under the fill strip the pressure is near ambient atmosphere or even positive head pressure so out gassing is suppressed only after initial fill….
virgin resin pocket under strip with trapped/solved gasses gets bypassed by ‘leap frogging’ resin after that because following resin flows around the static pocket of virgin thanks to flow strip medias lesser resistance.
Resin crossing fabrics between fill strips is constantly exposed to vacuum and fiber surface area at leading edge of moving flow front where gasses may nucleate on fiber surfaces and still have just a couple mm of path to escape out into dry fiber and exit to Vac pump…
2ndary strips fill slowly edge-across from constantly degassing resin filtered through fiber rather than virgin from bucket so no trapped pocket of virgin resin means no instant or even latent foamy stripes.


Possible corrections for process are:
-lower vacuum levels
-infuse beneath experimentally verified press/temp boiling point might suppress most nucleation and bubble will be smaller.
-I think even 20% the vac will get 80% the flow rate and 80% the compaction.
-pinch off resin inlet briefly upon filling just half the initial fill strips length and let vacuum recover there to effect resin while degass distance is only inch or so.
-give 1 min or 2 for wicking and degassing to make the patch under the fill strip more like the rest of the fiber pack will be later.
-reopen resin supply slowly (newly incoming resin should leapfrog around the initial wet spot like always and be degassed on the fly as front moves)
-perhaps slowly slide the fill strip over a bit to further flush out the virgin pocket into the flow stream and fiber pack.



i would really like to hear how things pan out

[otseg]

I am working on some bids now but when I get some time to digest your response, there are some interesting points to address. Here, by the way, is a finished Compmillennia 6.7 fast cat set up for tournament fishing. 0-30 mph in 3 seconds and just tops 59 mph. The bottom picture is 42 mph with one engine off and tilted up. 3.5 mpg @ 35 mph and 3mpg @ 40 mph. Efficiency K factor works out to be 10-5 pts more than a good V bottom. The rumor in town was that Fountain Powerboats was testing a secret military boat, at least until they were passed and they saw the Zuk's on the back. :-)

[gofish34]

Otseg graciously gave me the day off Friday so i could go fish The Big Rock Blue Marlin Tourney with a friend on his 50' Carolina Custom Sportfish - Miss Tammy. We didn't catch a winner but his mate did have a fishing license! When i got back Friday night - I hadn't had enough of the crystal coast yet so i had to go back Saturday. I'm (helping my Dad) in the process of building a 34' center console for our use - its got about 30 more to days to finish so i was able to grab the 22' Catamaran from the office to take out Saturday. I have ran many hours on the Pamlico River in this boat but we haven't gone out of Beaufort Inlet yet. Although it wasn't real snotty going out of there Saturday - it was typical for that inlet with 2-3' chop with occasional 4' roller coming in. I was totally amazed on how the boat ran and handled. To be honest i had never been in a catamaran in the ocean - it was unbelievably smooth. To be able to sit on the leaning post and drive in 3' chop @ 42 - 45 mph and not being beat to death is something you don't get to experience too often. My last boat i sold in order to build the 34' was 28' True World (Master Marine) center console with twin 225 Optimax - great boat - still miss it - but riding in 3 - 4 @ 40-45 you would get beat and be about 33 GPH - Saturday i was burning 12 GPH @ 40 - Now I have big decisions to make this fall when king fishing, with the group i normally fish with - they all fish really nice 28 - 34' center console production boats which I will have my custom that i can fish with them, but I'll be burning 25 GPH compared to 12 GPH and I'll probably be able to pass them if i take the catamaran. Oh well i still got a while to decide before cooler weather brings the big king bite.
I have decided I'll be making more trips to Atlantic Beach and few other areas in the coming weeks - if you would like to get a test ride on the catamaran just let us know and we will do our best to make that happen and as always we can always go out on the Pamlico any day - we gotta have excuses to get out of the office some.