infusing with contour balsa

Hi Groper - 200micron is 27Pa and yes this is much much lower than is possible in a usual VIP bag. I use a 25micron pump as well and the best bag I have done with that is 100Pa (99.9% vacuum). Using an accurate absolute guage measures the pump at the pump exit as 23micron. 500-1000Pa is a usual bag. If I try hard I can get 300Pa all the time. I'm reading a thesis at present that investigates 4 different closing strategies. The interesting one that produced the least porosity is to close the vacuum side but leave the resin side open. Will try this next bag and see what hapens. I usually have a process vac tube connected to the job and a scavenger circuit around the job (I use hessian tree tape or small rope) so I'll close the process vac and still run the scavenger vac but not close the resin port. Hmmmm

The article says that the laminate gets thicker as the resin front goes by and this is not my experience.

Peter

 

A well circulated modern practice in the composites industry, is to let the part fill, then clamp off the vacuum line. Leave the resin line open, but place the bucket about 1m lower than the job. This scavenges excess resin back to the resin bucket via siphon of approx the ideal head pressure to equalize the spring pressure again.. As it avoids scavenging via the vacuum pump, the voids also will not grow during the gel period. The system must have very good vacuum integrity or it will cause more problems than it solves compared to other methods.

"The laminate will get thicker as the resin front goes by" this agrees with what ive been saying... How else can you have a resin rich area near the inlet and higher Vf near the front? There must be difference in volume. Its the laminate spring pressure, or sometimes called "relaxation pressure" thats doing it...

 

Wow, this is fantastic reading, way above my pay grade though but will give many hours of interesting education for newbies like myself. As an update on my platform project i have added 3 x 8ft long stringers and g10 brackets for mounting a swim ladder and dive ladder underneath and it is on the boat. I am now in the process of making a mold for a filler piece that will essentially be a 6"x4"x12ft channel to the curve and angle of the transom that will fill the gap between the platform and transom created by the hydraulic raising and lowering arm geometry. Im still trying to decide if i want to infuse or contact mold it. I agree with Groper that not everything makes sense to infuse but i need the practice.
I am setting up the new pump and have a few questions, so far i have been using a single vacuum line around the perimeter of the part so only using a single vacuum line from the resin trap into the bag, the vacmobile site talks about breaking up the vacuum lines into segments and while i don't see a reason to do so on anything im doing at present i would like to include a provision for multiple vacuum ports in the lid of my new resin trap which is a paint pressure pot we will be modifying to suit. Vacmobiles resin trap lid has provision for 8 vac lines of various sizes. I have been using 1/2" irrigation hose from the resin trap to the bag and im curious what diameter is typical and why it would make any difference if it were 10mm,13mm or whatever but the suppliers of consumables do sell different sizes. The only reason i am using 1/2" is cheap hose and tee fittings that fit well with the spiral wrap im using for vacuum lines.
I guess im curious as to why one would ever use a different size vacuum line for any size project, i can see it on feed lines but unsure on vac lines, What size would typically be used for, say, a 45ft cat hull?
Another question is with regards to the use of an absolute vacuum gauge, where in the system do you hook it up? it looks like vacmobile hooks it up to the manifold on the side of the resin trap. I don't have one but would like to make provision for one later.
Regarding trying to eliminate leaks, i had an interesting conversation with a guy who was in the US navy on a nuclear sub and i believe he was involved in testing systems, they used a lot of vacuum and were constantly, like us, chasing leaks and he told me, its always the gauges, now im sure the navy uses the best gauges regardless of cost but he said they would put a plastic bag over the gauge and seal it with tacky tape and solve most of the leaks. I would never of thought of that this.

Steve.

 

Hi Steve = 1/2" irrigation hose is fine for all jobs. I use 6mm clear PVC for small jobs becuase its the biggest PVC tube that does not callapse under vacuum and it clamps very nicely and its soft. If you use irrigation hose in very hot weather it can be a problem. There is consumer grade LDPP pipe and commercial grade LDPP. The commercial grade is a little thicker then the domestic stuff. In hot weather the domestic stuff can callapse when the resin gets hot. If we are post curing the part under vacuum PP pipe is no good as it callapses in the oven so we use 12mm polyethylene pipe PE. Leaks are very strange things and it takes persistence and being very systematic to isolate them. I always start at the pump and put the absolute guage on it and work outwards or use the leak detector. The guage should be placed at the resin feed in line which is at the furthest place from the pump. If you put it at the resin trap you will always get a good reading. Cheers Peter s

If you put a guage (or gage in USA) in a bag you isolate it from the atmosphere and if its a vacuum system the gauge will stop working as the pressure on each side of the bourdon will be the same?

 

Steve; I also use an old paint pressure pot for the resin trap, have one 1/2" inlet and two 4mm. Have only used 13mm commercial LDPP tube for vaccum line once, I can get the 4mm micro irrigation tube and connectors much cheaper than the 13mm. For vaccumm lines this is big enough, for a long time now I havent had any resin in the pot and can reuse the vaccum lines. I use the 1/2" connection as the primary, stepping this down with a short section of 8mm CVT and then the 4mm LDPE. The clear tube alows me to see if any resin has made it that far. From this one connection I will branch out using T's to as many vaccum points the infusion requires. When you start doing more complex shapes breaking up the vaccum ito segments allows you to control the resin flow front by shutting off sections already filled. The two 4mm connections are spare vaccum sources so that if there is a lockout you can put the vaccum into that location to pull the resin into the dry spot.

Peter; you misunderstood me, Groper understood what I am advocating.
1. use the maximum vacuum available ie. full pump capacity to dry out your reiforcement, have maximum compression and have minimum air in the job.
2. minimise mixed in air in the resin
3. infuse at same maximum vaccum
4. maintain maximum and constant vaccum

Thats it for a lot of the jobs, but for times when you know that you had more mixed in air in the resin than you would have liked then reducing the vaccum ie. increasing the absolute pressure in the bag is a means of minimising micro voids in your laminate.
The only trade off here is increase in the laminate resin weight %.

I think this and the mechanism of why this helps is what we disagree on.

I will end up repeating my self here but will try and explain my reasoning again.

You want to infuse at max vaccum / minimum absolute pressure so that you start with minimum air in the laminate and to get the maximum benefit of the ideal gass law to remove as much of the dissolved and mixed in air in the resin. This subjects the resin to the maximum posible pressure change, this leads to maximum expansion of microbubles as well as the larger mixed in air bubbles. This is mostly taking place at the resin front at the surf zone and is scavanged out by the pump.
Further back from the surfzone you no longer have the same absolute pressure it has increased ie. less vaccum. Now we have a pressure gradient, minimum absolute at the dry reinfocement ahead of the resin and highest absolute at the resin inlet.
This means that the micro and larger bubbles dont get to expand as much at the inlet, remember its a agradient between the resin front and the inlet. While there is resin flow even these bubbles away from the surf zone get scavanged out with over the top infusion media.
The other thing that has occured due to this pressure gradient is that the externall compression presure has decreased by the same amount that the absolute internall pressure increased. This means that the reinfocement at the inlet has decompressed the most and has got thicker, resin has taken up this volume. Temporaraly some of this volume is taken up by the expanding bubbles but these are being scavanged out so dont matter at this stage.
So now we have 3 gradients, internall absolute pressure, externall clamping pressure and laminate resin wt% fraction. At this stage I would also suggest that its only at the surfzone that we have scavanged out all gases (dissolved, micro and visible bubbles) and at the resin inlet zone perhaps just the visible.

OK now the infusion is complete and in most cases we just close the resin inlets and leave the vacuum open. What hapens now with time is the 3 gradints slowly deminish.
Resin rich area near the inlets slowly migrates out, absolute pressure goes down producing greater externall compression, laminate gets thiner and resin wt% fraction goes down.
The undesired effect is that as the absolute pressure decreases at the inlet zone and remember its an actual gradient from the resin front to the inlet more seeding of disolved gasses and expansion of micro bubbles takes place untill a new equilibrium takes place.
Unfortunately microbubbles and or visible bubbles below a certain size do not escape the laminate.

Well to minimize this if the vaccum is reduced to say 85% ie absolute pressure increased before or just after the resin inlet is shutt off and maintained at a new steady state then the micro bubbles will be greatly reduced. This is because the resin inlet zone will not be subjected to full vacuum with time as in the above case.
At this stage there is enough resin in the system to redistribute itself from the resin rich inlet zone outwards into the resin front. There is no net volume change taking place inside the bag so no concern about decompressing the stack.

Very late here now so will add to this later.

Cheers
Andrew

 

Hi Andrew - I agree with what you say, well explained thank you. I like the term clamping pressure and will add that to my descriptions list! But the external pressure (clamping pressure) is not a gradient, and I would not describe the weight fraction as a gradient its an outcome of the process. But it is a process variable. But your general description is good, thanks for being patient AK & Groper. My aim is to develop the maths for the process so I have to be very rigid in descriptions. I've read several research papers that measure the laminate pressure during the infusion and they have conflicting conclusions. They generally use oil or glycerine for the research (SO THEY DON'T BOG UP EXPENSIVE SENSORS lol) and I think this excludes our outgassing problem. It does produce good data on pressure profiles but doesn't help us to cure our bubbles problem! Plus they are generally small 300mmx1000mm panels so are easy to do. I intend to use clear PVC tube as manometers on one of the next jobs to use the resin as a pressure gauge in various parts of the job. Unfortunately the job I'm working on at the moment wants results not research so I haven't been able to "play" with anything just tell the fabricators what I want and leave it to them. Its been a two year program and we have just passed the endurance test and ultimate load tests this week so I'm a happy camper. Maybe I'll be able to do some more "research" now on these bubbles. But to summarise there are a couple of ways to close off an infusion and each one creates a different pressure profile in the job. The trick is to create an advantaguous profile not a disadvantaged one ie one that does not decreases the absolute pressure quickly encouraging outgassing. This is typically what clamping off the resin line does. It is also a dynamic effect (as the resin is moving) so I have to bring bernoulli's eqn (bernie was quite bright) into it so it hasn't been easy. Maybe the easiest approach at the moment is to clamp off the resin slowly (over say 1 or 2 mins) so that the pressure does not spike and the bag has time to adapt. Its like turning off a water tap fast and getting water hammer (thats exactly whats happening I think as its an elastic system!! eureka) so good to talk and think... Peter s

 

Hi Peter, I am enjoying the constructive discussion also.
To answer one of your earlier questions, no I dont have resin in the vaccum tube when I increase the absolute bag pressure.

At the moment only have time to expand on minimization of micro voids, will come back to what I describe as resin wt% gradient and clamping pressure gradient.

Another way of describing of what is going on is to think of the infusion period as a resin degassing period too as that is what is occuring. Now lets put some numbers to the absolute pressure inside the bag, say 300Pa before we open the resin inlet. This will give us 300Pa in front of the resin front and say 30,300Pa at the bag resin entry point so a gradient of 30,000Pa. Or in vaccum% terms 100% and 70%. We can now say that we degased the resin at the surf zone with 100% vaccum but the resin at the inlet only at 70%.

When the infusion is completed our degassing period 1 has also finished and now we clamp the resin inlet. This is when degassing period 2 begins as now the absolute pressure of 30,300Pa at the resin entry point starts to decrease. As this decreases to 20,0000 then 10,000Pa more seeding is taking place and forming microbubles untill a new equlibrium is reached say at 5,000Pa

But if at the end of period 1 we increase the absolute pressure to 15,000Pa then the new equlibrium is established at this pressure and not the 5,000Pa. This means less bubbles formed which is what I observe in real laminates.

 

I think that the pressure profile across a part is driven by the dynamic pressure difference (V squared/2g as most parts are flat therefore no static pressure difference) and the friction. When the resin line is clamped the pressure spikes and then takes a while to settle. I'll write up a calculation sheet and publish here soon for mud to be thrown at it. have to figure out how to include the springback pressure now. Cheers Peter S

 

There must be a pressure gradient across a part.

As the vacuum has much dissolved air at the surf zone ( I like that terminology ) then the pressure will be very low, close to the dry zones. As you move further back from the surf zone, there is mobile bubbles and more resin, moving further back still, less and less bubbles, until there is only air in solution ( invisible). Once there is virtually no air bubbles, and therefore almost zero elasticity, you have the ambient pressure acting on both sides of the bag, not forgetting the spring pressure in addition, on the inside. This spring pressure will temporally expand the bubbles in solution as they go from 1atm, to ~850mbar entering the bag.

So a pressure gradient seems highly likely.

On the whole, the entire dynamics is very complex once you factor viscosity resistance, pressures, liquid/gas mixtures, etc etc...

Good luck with your maths peter, many very intelligent people have attempted to fully explain the process entirely, with absolute certainty, but thus far no one has succeeded i believe. A better avenue of persuit seems to be controlling the process further, to eliminate the parts of the process that are known to cause problems and produce more consistent , higher quality laminates ...

 

Peter; agree with you that for dissolved gasses to outgass at normall temperatures you need >95% vacuum. So what I am talking about is the mixed in air microbubbles as large ones are scavanged out while there is resin flow.

OK lets get back to my description of clamping pressure and resin wt% fraction gradients.
I say that these exist during the infusion time and for some time after the resin inlets are closed.
I think that we agree that the absolute pressure at the dry laminate in front of the resin front is the lowest and highest at the resin inlet zone. If this is the case then at standard atmospheric pressure the clamping pressure in this zone is ~101,300 - 300 = 101,000Pa using the example in my post #112.
At the inlet zone 101,300-30,300=71,000Pa, ie a clamping pressure gradient of 30,000Pa.
This is reflected in the resin wt% fraction, highest at the resin front and lowest at the inlet zone. And yes this is not fixed, gradients diminish with time after the inlet is closed. We see this as resin flow for a while after the inlets are closed. Only then does the resin wt% become a fixed variable of the process.

Reflecting back on my early tests and I am looking at the sample now. I did a trial to see how far the resin would flow just through glass reinfocement alone, it shows what I am talking about. The laminate feels thicker and stiffer at the inlet end then it does at the resin front end. Also no visible voids to the naked eye at the inlet end and progresively voids increase towards the resin front.

 

Hi AK and Groper -

 

VIP Theory

Hi AK & Groper - Lets agree to the attached set up and I'll try to describe it using physics. I call this a basis VIP setup. It does not have a vacuum scavenging circuit. Its a standard test panel for my purposes. 600x600 is a half roll cut. 600x600 is what we use to send to test lab for tests. Typically it takes <15mins to fill depends on the media used. So to start we need to agree what is the starting pressure for the job? I usually start the bag at <1000Pa abs (99% vac). Lets agree to a starting point and I'll write it up. Please make any further comments or adjustments required to the sketch so we all start at the start... or we can use any setup you like. If the maths stand up it should stand up on any setup... I also use a 300mm wide x2m long to figure the flow rate vs distance and the variables for Darcys eqn... Cheers Peter s

1atm=101kPa = 14.7psi = 1000mbar

 

Not sure about your fibre volume fraction... i get around 65% using dbias e-glass... Although this might be a product of other parts of my process...

I calculate the resin quantity required at 65% Vf, then clamp the resin line once the bucket is empty. The part is not fully wetout at this point, it continues to pull resin from teh resin rich areas near the inlet, then stops once the part is fully wetout and the peel ply resin brake is saturated about half way to the vac line. - No resin gets into the vac line.

Im reasonably confident ive had nearly 70% Vf in some instances - all multiaxial laminates, no wovens of course... But i always use a more conservative estimate of 65% so i have a bit of extra resin on hand to make sure it completely wets out.

EDIT: - OOPS, i referring to a 65-70% weight fraction, not volume fraction - carry on

 

VIP Theory

Hi Groper - Very unlikely you are acheiving 70% Vf. This would mean you only have <20% by weight resin in the laminate. With VIP glass laminates most people acheive Rw=35%. Using advanced vip I have built laminates at 20-25% resin by weight. Average for a good bag is Rw=30%. I've had lots of laminates tested over the years and most people I train get 30%Rw(50% Vf) for glass. Boeing have a program to acheive 15% and they VIP the part then put it in an autoclave and squeeze out the require resin to get this low. Prepreg acheives around 35% regularly. Its supplied in various resin ratios and by using a dry layer and bleeding the fabricator can get the Rw down. But not much point to make laminates under Rw=30% as your interlaminar shear strength drops hugely. At Rw=35% with a good resin you get around 50Mpa then at <30% it drops to 35Mpa which is a bit low. Aerospace resins are extremly tough and get shear in order of 100Mpa at Rw=30%. Cheers Peter s

 

just saw the edit - peter

There is a guy in USA advocating double bagging to get to 70% Vf but his claims have not yet been able to be replicated by anybody. Buy the way CSM gets to Rw=30% as well. makes a nice strong laminate cheap if you can cope with the 2x stacking factor. I did a 50mm chainplate once and the CSM stack was over 100mm high to start with.

Buy shutting the resin early you are letting the bag reprofile and redistribute pressure before the fill end occurs this is a very good strategy... So what starting pressure do I do the maths for??