Infusion related - will this idea work?

[Wynand N]

New to resin infusion and still tinkering with some test runs, I'm always on the think how to speed up things before doing my first hull.

I now have this idea to boost the resin flow the first minute or two to get max coverage sooner before things slows down or start gelling. The problem is will it work before spending money to pull it off.

The idea is to take an old 49kg jumbo gas cylinder, weld two sockets onto the side (fit a vacuum gauge on existing outlet) of the tank - one to the resin trap, the other to the vacuum pump. On the pump side socket, fit a T piece, valves so one cab bypass the cylinder by turning a valve and suck directly from the resin trap to the envelope.

In principle, according to my distorted view of things, one pulls the big cylinder to a full vacuum, close the valve to keep it in that state, switch to the envelope and pull in to a vacuum. When all is ready to open the resin feed lines, switch over to the cylinder under vacuum and open the resin feeds.
I believe the volume of the cylinder vacuum will be so potent in sweeping the vacuum in the bag when the resin flows in, the process should happens incredibly fast until the vacuum in the cylinder drops a bit and then just bypass the cylinder directly to the pump to maintain the vacuum for final bit of infusion until the resin set.

Will this speed up things as I believe it will or am I missing something:?:

[Eric Sponberg]

It should. This is the principle of the plenum chamber. The larger the plenum chamber (your 49kg gas tank) the more vacuum power you have. I have run vacuum molding (prior to the resin infusion common practice) with a really large plenum chamber, a plywood box 4' x 4' x 8', covered completely in fiberglass and reinforced heavily on the inside to keep atmospheric pressure from collapsing it. We would draw our initial vacuum off the plenum chamber to about 13 psi vacuum, and then open the valve to the vacuum bag over the laminate--sucked down instantly--ZAP! After we inspected the back for leaks to make sure it was holding, we would stabilize the plenum chamber to about the 13 psi vacuum, and then shut off the vacuum pump and close the valve between it and the plenum chamber. It all worked great. So with practice, and you may have to tweak it here or there, your idea will work.

Eric

[KnottyBuoyz]

Yup, it'll work. I have a vac reservoir in my setup. Helps a lot with pump cycling as well. Larger resin lines with a broader distribution system and flow media will speed things along too. There's some cool vids on Youtube of Henny infusing his F39 hull that'll show you how the larger projects work.

http://www.youtube.com/watch?v=JITbV-_C60w

www.fram.nl (http://www.fram.nl)

[AndrewK]

Wynand,
No I dont think so, large vacuum chamber will only draw down the bag quickly. this is why you can do large infusions with small pumps provided you are happy to wait a long time to pull down the vacuum on a large job or to use a vacuum cleaner for the initial pull down.
What determines the speed of infusion is:-
1. resin viscosity
2. vacuum, bigger vacuum means bigger driving force. It is the pressure difference between inside and outside the bag that is doing the work, this is where locations at high altitudes lose out to some degree.
3.reinforcement stack resistance, more open spaces in the reinforcement means less resistance. Different types of reinforcements compress to different degrees. Replace your chopped strand mat with CFM for starters.

Regards
Andrew

[waikikin]

I got a vacuum resevoir too, we call it the vacuum "PIG" cos it kinda looks like one, I might even paint it pink one day when I got time.

[Wynand N]

bigger vacuum means bigger driving force. It is the pressure difference between inside and outside the bag that is doing the work, this is where locations at high altitudes lose out to some degree.

That is my point, bigger vacuum means bigger driving force...My view is that when you open the resin feed, the volume of resin entering the vacuum envelope, needs to be swept out of the bag by the vacuum pump to maintain the vacuum - and usually smaller vacuum pumps do not sweep a great volume, in my case I use a 6 CFM unit.
Having that bulk of vacuum power sitting in the storage cylinder, one can sweep a lot of resin into the envelope quick quick before the vacuum has to be sustained by the pump. Albeit my point of view on this

[KnottyBuoyz]

I use a 6 CFM unit.
6 cubic feet is a lot of resin to pull through a layup! :D

My 5 CFM pump can easily achiever 29"HG which is pretty near perfect IMHO. So anyways once you get the vac down to that level it doesn't matter if it's in the layed up panel or the vac reservoir it'll have the same atmospheric pressure to drive the resin into the layup regardless. The advantage of the reservoir is the ability to draw down a large layup fairly quickly.

So in essence you could have a 300HP diesel driven commercial vacuum pump producing 29"HG vacuum or a 1/3HP 5 CFM pump producing 29"HG vacuum and you'd still have the exact same atmospheric pressure driving the resin into your layup.

I prefer to spend more time & effort to ensure a perfect seal on my layups than worrying about how many cubic feet of air my pump can move.

[AndrewK]

Wynand,
What you are proposing is bigger volume of the same vacuum your pump is capable of producing, not bigger vacuum.
I think previously you mention that your pump produces approx 86% vacuum with a 6cfm capacity. 6cfm is a huge amount of resin to be sucking in. 86% vacuum is OK but not great, with vacuum infusion the greatest driving force you can achieve is to have 100% vacuum.
You can further optimize your system with lower viscosity resin and more easily infuseable reinforcements before going to the higher vacuum option.
Most of us locate the vacuum gauge on the catch pot for convenience and protection from resin, but on large jobs we should checking to see what is happening in the bag.

Andrew

[Wynand N]

Wynand,
86% vacuum is OK but not great, with vacuum infusion the greatest driving force you can achieve is to have 100% vacuum.

And herein lies the problem; my workshops are about 1500m / 5000ft above sea level and the barometric pressure about 24.9 and that is quite a bit off the 29.92 for an absolute vac at sea level.
I am thus stuck with a vacuum in the mid 80 percent about - one can not take out more than you put in - hence the bigger volume of vacuum to speed things along.

You mentioned CFM - do you have a link to this as local distributors do not know what this is, so do I.

[Rick Willoughby]

I expect the cyclinder is about 100 litres in volume. If the space around the bagging is also 100 litres (that is a lot of space) then the pressure will be only half vacuum after the initial inrush. So if 50kPa is enough then this will work.

By comparison with Eric's 1.2m x 1.2m x 2.4m (3456 litres) your cyclinder is puny. If it was pulling down 100 litres to start with the vacuum would hardly change.

So it gets down to the relative size of your cylinder and the spacing around the bagging. If you did an initial pull down with high volume low vacuum like a vacuum cleaner then this would give you a good start. The actual space in the cloth is miniscule - just a few litres. You could valve off a feed to the vacuum cleaner to get an initial pull down then close this off. Pull the full vacuum in the cylinder and the job then open the resin valve. If you have say 10 litres of resin then the vacuum will still be up around 80kPa after infusion. This would be much faster than just pulling with a little pump after opening the resin valve.

Rick W

[Wynand N]

In principle, according to my distorted view of things, one pulls the big cylinder to a full vacuum, close the valve to keep it in that state, switch to the envelope and pull in to a vacuum. When all is ready to open the resin feed lines, switch over to the cylinder under vacuum and open the resin feeds.


Rick, the vacuum in the cylinder is not to pull the bag down, but to have a large boost of vacuum on hand when the resin feeds are opened after the bag is pulled down and the vacuum established, and then switch to the cylinder. I believe (hopefully) that this initial rush (volume) of vacuum pulling power will give one a head start...

This should have a twofold purpose as well - as the pump get hot keeping the vacuum intact the cylinder can act as an accumulator and not be sensitive to little losses as the pump.

Since there are different views on this, I think I must take the plunge, get an old cylinder and do the mod. I will layup two identical test pieces and let it rip - one with the "storage cylinder" fitted and one without. The results will tell the true story.

Any predictions on the outcome:?: :D

[Rick Willoughby]

If you pull the whole system down to good vacuum, including the cylinder, before you start the resin feed it will make a big difference unless the volume of resin to infuse is of similar size to your cylinder or there are lots of leaks.

You cold also apply a little positive pressure on the resin feed drum to overcome flow losses in the feed tubes.

Using the cylinder will be far better than just having a little pump.

I am very interested in the outcome because I have not used infusion although I have used vacuum bagging.

Rick W

[wet feet]

The best vacuum system I ever saw featured a large tank and both the tank and the cylinder were outside the workshop.This was great from the point of view of finding leaks as there was no noise coming from the pump to confuse things.The system was built to operate along the same principles as an air compressor-in reverse-with limit switches to switch the vacuum pump on or off as needed.This helped to prolong the life of the pump and the large size of the tank reduced the delay between starting the operation and achieving full vacuum.
The CFM referred to in an earlier post is another compressor like reference Cubic Feet/Minute.

[AndrewK]

Wynand
If you already have a spare cylinder then do the experiment and find out for your self.
Although I do not think it will help you I agree it is useful by being able to turn the pump off when checking for small leaks.
Think of this in another way say water analogy. Say you have a 1000L tank that is only 1m high and another that is 100L but 2m high, if you drill same size hole at the bottom of each which will give you the greatest pressure? The smaller one will have twice the pressure = twice the driving force but will only last one tenth the time. Same thing with vacuum, its the ultimate pressure differential that gives the driving force not the volume behind it.
Your 6cfm = 28L/min that is a lot of resin to be sucking in.
Do not pressurerise the resin container, it will only flood the area around the resin inlets.
You mentioned your hull is approx 8m, say you have approx 30m2 and 5mm average laminate thickness. The volume of laminate is 150L and volume fraction of resin is 60%, this means you need 90L of resin. Your pump can do this in 3 minutes.
But in reality it will not happen this quick because of the resistance the reinforcement presents.
With vacuum infusion this can be overcome in many ways, you need to determine what is the most economical for you.
Order is from simplest and cheapest:-
1. long gel time resin
2. low viscosity resin
3. do not use CSM, CFM of equal weight is 3 - 4 times faster to infuse
4. speed optimised infusion strategy;
bigger bore resin inlet & feed lines
closer spacing of lines
faster over the top medium
fish bone versus sequential plumbing
5. interlayer imbedded infusion mediums between reinforcement eg lantor infusion coremat, enkafusion mesh.
6. use infusion specific stitched reinforcements in place of woven roving


Google CFM, continuous fiber mat or continuous filament mat
Google Enkafusion
Google Lantor Soric XF grade mat
Also have a look at http://www.vectorply.com/ they make infusion friendly reinforcements some have CFM backing. Also you can download free laminate prediction excel program from there.

Lance Brown may also be able to help, he acts as a broker so while he is based in Australia he can arrange delivery of product directly from manufacturers to you but there are minimum orders so not much good for one off builders. At times he has very good deals on surplus stock if you are in the position to take advantage of these so get on his mailing list if you wish.
mail@LBIE.com http://www.LBIE.com

Regards
Andrew

[Fram]

While it is nice to have a big resin trap, you don’t speed up the process with a bigger tank. The most important influence on speed is flow length. Want to speed up things, just make a smart infusion strategy. As Andrew said, there is a big difference between the fishbone and the parallel plumbing. The first doing big infusions within one hour, even single handed, the second requires much more time and operators (well, at least that looks quite professional ;) )

It would be nice to use just the right amount of resin so that the resin trap remains empty. For serial work one knows after some time when to stop the resin supply. For one of projects this requires a lot of nerves. Better using some more resin for peace of mind.

But the problem of excess resin in the resin trap is heat built up. Apart from my normal resin trap (a 400mm pvc tube with a 12 ltr bucket inside) I use some 125mm pvc pipes in parallel. These are easily removed out of the vacuum system when time is ready (= pipe getting warm)
When infusing a 40’hull I have my two 400x400 resin traps and two 125mm pipes in parallel (length what is available on that moment). Thus having a “big” tank in big infusions, but not because of filling speed.

Henny van Oortmarssen
www.fram.nl

[Wynand N]

Got a 115 liter cylinder today and will start the modification as soon as I have time on hand and hopefully do a test this afternoon to see if the "vacuum booster" will work or not.

Andrew, the polyester resin I use is a NCS 301 PA LSE, an isopthalic NPG (new penthol glycol) low viscosity type - much thinner than GP polyester resin and has high waterproof and chemical resistance properties.

The viscosity is fine as is evident in the layups I had experimented on - the last drystack consisted of 3 x 450g CSM, 2 x 2mm Coremat and 2 x 610g WR and it went through that reasonable well but my enemy is the gel time. In my part of the woods the heat is in the high thirties/early forties C in the workshop and a less than perfect vacuum due to high altitude which I must do battle with.

However, I spoke to my fibreglass pimp this morning and was told he has a slow gel catalyst that in effect doubles the gel time. He has all the potions to thin out my resin should I need that, but if the gel time is doubled with the new catalyst he is sending, I would prefer to keep the resin "intact" for obvious reasons.

[Wynand N]

And my question has been answered;).

I has set up the test bench with two identical sized drystacks consisting of 2 x 450g CSM, 2 x 610g WR, 1 x 2mm CoreMat and surface tissue.

The drystack with pump only was hooked up to the 6cfm JB Ellimanator and pulled a vacuum of 87kpa.

The cylinder driven unit was hooked to the old Edwards pump and due to the age of the pump, only got an 82kpa vacuum in the cylinder.

Each stack had its own vacuum pipe to its own unit and individual resin feed lines. The resin was catalyst in a single container and then split in two smaller containers to keep things even. At this stage I must confess that I had a leak on the cylinder assisted stack we could not get closed that may have a slight influence on the result.

The photos tell the story. The feed lines were opened at exactly the same time and at the end it was a close call with the cylinder ahead by a very small margin. Cigars to some and well, it was worth a try to myself and others. As the saying goes; nothing ventured, nothing gained :idea:

Was all this a waste of time and greenbacks?
In short, no. This experiment showed me that the cylinder in a full state of vacuum, can be a valuable backup unit in the event of losing a pump, or electricity break - just switch a valve and one is safe for quite a while to get another pump started, or the generator going.
I am going to build a compact vacuum station on a steel platform with caster wheels so it is movable - will have the JB and the Edward pumps fitted permanently to a manifold system that is hooked to the "storage cylinder" and the resin trap fitted as well. From the system just one pipe going out that pulls the vacuum.
In this setup I can change to any pump at any moment, or just suck with the cylinder (excellent idea for smaller parts - no noise) or all units and cylinder together.
Will post photo of the setup when sorted out and build. The little Gast pump will be kept as a backup to the main system.

[MarboMan]

"Each stack had its own vacuum pipe to its own unit and individual resin feed lines."

You can get a "bag" fitting from Air Tech and try pulling the vacuum from the center of the stack and have resin feeds at each end and put a resin trap before the pump so it doesn't inhale resin.

[TeddyDiver]

Nice work W.. :D

[AndrewK]

Wynand,
When new to infusion no experiment is a waste of time, as discussed the large tank will be good for the initial pull down, finding leaks with no pump noise and as you say good reserve of vacuum if you lose power.
The slower hardener will make a big difference and hopefully at no extra cost. From here on all other improvements to speed things up will cost more, directly in materials or through wasting more resin or both eg more of and bigger bore tubing.
Is the new hardener the same Fanie used for his beams?
Do you know what the resin viscosity is? can your pimp find out? It would be nice to know what you are dealing with.
I imagine even a 5% addition of styrene would give a large reduction in viscosity.

Have you determined how much the coremat is being compressed, my thinking is that it may not be adding much thickness to your laminate and therefore waste of material.
For comparison the data sheet for the infusion grade states that it loses 10% at 80% vacuum.

Another experiment I suggest you do is to make your maximum thickness laminate to make sure there are no excessive exotherm issues.

Also now that you have a number of test laminates it would be good to do a simple test and see how much stiffness is being lost with infusion compared to open lay up.
Say cut a 50mm x 600 strip of each laminate and clamp to a table so that you have 500mm overhanging. Hang weights on the edge and measure displacement.

Cheers
Andrew

[Wynand N]

Andrew,
Fanie is using a different resin altogether, but from the same manufacturer. I traveled today to my fibreglass today and he gave me a liter styrene to experiment with and recommends I start with 2% and work my way up to a max of 7%.
I also got a new catalyst from him that doubles the gel time and I did a test with a little resin with the standard hardener and the new stuff and the gel time is stretched to nearly double - that may help considerably.

The data of the resin I use as per spec sheet is;
Viscosity @ 25°C, mPa.s 200 - 300

[AndrewK]

Wynand,
200 - 300 mPaS @25'C is good, in your case the lower it is the better.
At the moment while your workshop is at 35'C actual viscosity would be a lot lower.
A few more suggestions for you, when doing the trial always record temperature, vacuum (in the bag), plumbing layout and distance resin travels per minute. You will find this very valuable when planning new infusions.
There is a big difference between 200 & 300 mPas, if your resin supply is inconsistent and if you have a ford cup for measuring paint viscosity it may pay to check what the flow time is for your current batch of resin. This can be then used as a reference point for future batches, and if they are higher in viscosity you can make adjustments for this.
So far you have also done your trials with the inlet at one end vacuum at opposite end, do your next few with the feed line in the middle and vacuum lines at either end so the resin travels in both directions. On larger jobs you will be using a combination of both so you need to know the behavior of both feed systems.

Cheers
Andrew

[jim lee]

Wynand, If I read this right, your infusing chop strand mat and woven roving? If this is so, a better experiment would be trying a stack of biax with infusion core.

I -think- that with the material you are using, the flow rate is so slow you will not see the benefit to your tank idea. With the correct materials I'm betting your going to see a much faster flow rate.

Good luck!

-jim lee