infusing with contour balsa

No core Steve,just playing with finish and final thickness.With all my tests so far the one ones that have gone slow turned out the best.
I'm also only infusing at 25Hg as that's all I can get.

 

Groper, I think your probably right on the money, thanks for the vacmobiles link, its going to take a while to absorb it all but its invaluable info. Jeff, you've got to try this, after 40 years in the boat building industry im having more fun than I have in years.

Steve.

 

Good discussion

Sorry I got to it late. Steve, I also own the Amprobe ULD-300 and I believe you are correct it does not perform. There is a discussion on the Composites Central forum a while back discussing leak detectors and no one had tried the Amprobe but reliable members did like other products, you should look into it.

Vinylester gave me fits until I cut the vacuum back, as said it will boil. I use epoxy now but even had trouble with one of them at full vacuum.

To your original concern about infusing balsa, it will have a moisture percentage compared to foam which with enough vacuum and heat over time could be eliminated. The balsa I have in the shop currently (Diab, I believe and a few years old) has about a 25mm x 50mm knife cut, but the ends of perhaps 1meter panels are 50mm x 50mm on both ends? Maybe they have changed this, but even so the cuts vary from tight to loose. I bought the stuff and never used it because I thought race tracking and dry spots would be problems. That said, a good flow medium/reinforcement on top and bottom can work.

To Gropper, my thoughts exactly has anyone wrapped spiral wrap or irrigation hose with Tyfec? Cheap MTI makes sense to me. I would love to hear about anyone experimenting.



Jim

 

Steve, I've done a coupla little "experiment" things but pretty keen to learn, I just need to spend the time & have a scheme to do some infusion on some daggerboards I'll be needing to make. It's great that your having a go & that blokes like Groper are sharing what they've learnt.
All the best from Jeff.

PS: I like to have fun too.......... maybe that means infusion!

 

Steve, looking at your photos I think its air that is your problem. If you had no leaks and say that the glass was translusent after the initial infusion then its mixed in and dissolved air.
With my very limited use of ve resin I still say its not styrene boiling off, you can prove this by placing some resin into a small clear container and then placing this into a larger clear container. This can be a glass jar, it will take full vacuum, bring this to max vacuum you have. Initialy you will see some bubbles this is disolved air, this will clear and there should be no further bubbles.
Was it due to csm binder? dont think so, you would be having the same problem with wet lamination. Infusion will only improve the wetting out of dificult reinforcements and not the other way around.
Was it moisture in the core or glass? no this produces a milky looking laminate.

What I think is happening is that you have mixed in air in your resin, when the resin is introduced into the job a lot of this is transported out. You will observe lots of bubbles at the resin front. Its most obvious there as that is where you have your max vacuum causing greatest expansion of mixed in and dissolved air. Away from the resin front the vaccum would have decresed reaching a minimum at the resin inlet and you dont get the same expansion of these gases. When the Infusion is finished the vacuum slowly reastablishes across the infused laminate, this is when the tiny bubble start to grow in size and turn what looked like a great infusion to something lesser.

What can be done;
1. do not start an infusion untill there are no leaks at all, I do not bother with drop tests any longer. Instead look for air output from my pump, there should ne nil.
2. de gass your resin first, this is a luxury only afforded with long gel time resins.
3. burp your resin lines first to get the air out.
4. slow the infusion down, use slower medium and smaller resin inlets and feed tubes. This maintains a higher vacuum into the job during the infusion which helps pull out more of the mixed in gasses.
5. last option, only do this if you know you have excess resin in the laminate and can run your pump at reduced vaccum. Very soon after clamping off your resin inlets before the vaccumm has redistributed it self into the job reduce the vaccum. The reduced vaccum means that any mixed in gas left in the job at this point will expand less.
I do this at times as I have a small diphragm pump that has ~85% vaccum capacity, I switch this for the initial two stage pump which gives almost perfect vaccum. The 15% reduction is OK for the amount of excess resin I have in the job straight after clamping my resin inlets.

Keep experimenting and learning, good infusion provides completely translucent laminates, be it 1mm or 20mm.

 

Andrew, for the next part im going to make just a couple of changes, i will draw a full vacuum which is about 29hg at my altitude, try to eliminate all leaks and then hold it for about an hour, then reduce it down to about 23hg which is about 80% before introducing the resin, then I will leave the vacuum at that until the part has reached full exotherm and starts cooling down. Im not sure I can do much about air in the resin other than not stirring too vigorously, im not using a power mixer, just a stick. I don't think its practical to vacuum degass at this point. Back in the 90s I used to manufacture snowboards and they were press molded and there were very difficult to problem solve, much more so than anything ive experienced in boatbuilding, and one thing I learned is that you don't want to try too many things at once or you never figure out what the problem was.

Steve.

 

Steve, do not reduce the vaccum untill the infusion is finished, you want perfect vacuum at this stage. Also dont go below 80% after infusion as you run the risk of not having enough excess resin in the job to keep the reinfocement filled as it decompresses. If you are using an oil vane pump first check that your pump can be run at reduced vaccum. For mixing I can recommend a squirrel mixer these are very good and will not suck air in unless you lift it too high. They are most efficient I have come across, I use the small ~50mm diam plastic one and only with one fan. They are hard to get here but are US made so should be more common for you, if the local paint store does not have them get one from ebay.

 

Andrew, he is infusing with VE and PE resins.... full vacuum will boil the styrene at the resin front...

 

I halped a friend do an infusion job with VE and that was not the case.

 

Yes, everyone has been telling me to use lower vacuum with VE and PE. I also don't think i have significant moisture in the materials, in fact I cant get any reading with my moisture meter at all in the balsa, it is very dry here in the winter. Ill pick up one of those squirrel cage mixers and give it a try.

Steve.

 

Well, we infused the swim platform yesterday and it went very well, the resin fronts progressed very evenly and it took 35 minutes to completely wet out. I used 1" coosa in the two areas where it will be mounted to the hydraulic arms so I drilled on 38mm centers and joined them on the tool side with 3mmx3mm kerfs, lesson learned here is I would skip the kerfs next time as the front on the tool side moved ahead of the top. About the only other minor issue is we did get some bridging in the radiused transition from the horizontal to vertical surfaces even though we did our best to work the glass into the corners under low vacuum during setup, we did use spray adhesive sparingly on the vertical surfaces to keep things in place, anyway, no big deal, just a bit rich on the radiuses. We sucked it down at full vacuum for an hour and then backed down to 23hg which is about 75% for us. The platform is 74ft2 and has a total of 54.3 lbs of fiber, 31.3 lbs of 1" contour balsa core, 20.3 lbs of coosa and consumed 64 lbs of VE resin which looks like very good fiber fraction but I still cant separate it out from the core uptake. I just have not been able to get that number from anywhere online so I put a call in to Baltek (I don't know why I didn't do that long ago) last week and they are going to get me a number but I need to go back and do another couple of test panels separating out the core and fiber as this is the missing piece of the puzzle for both estimating how much resin to mix and final fiber fraction. As it turns out the regional sales rep for Baltek supplied me with all the materials for a foam cored uldb i designed and built back in 1987 when he was with a different company and had actually visited during construction so he has been at it a long time and will be a great resource, some interesting things he told me was that all the big production builders are gelcoating and hand laying a heavy, at least 2 oz skin coat before infusing for cosmetic reasons as well as so they can walk around in the mold. Another interesting thing was that their 5lb density balsa consumes less resin than the 10lb.
Anyway I will have my son post some photos of the platform in the next few days.

Steve.

 

Hi Andrew & Groper - Welcome to 2014 Steve W - welcome to the world of infusion. I'll comment on a couple of things. 1) I have done quite a bit with balsa and plain balsa is a dream to use for infusion. Resin flows through the pores to the other side and resin take up is acceptable. Steve if you have some perforated film next time you do a test place a small piece of balsa in some perf film (weigh the balsa first) then infuse. Peel of perf and reweigh. This will give you the take up weight (same as with any other core). Same as contoured although the adhesive used to stick on the scrim can sometimes inhibit the resin flowing through. 2) I disagree with reducing vacuum at end of job. This means the bag has to get bigger so there must be either extra resin or outgassing to allow the increase. So ideally you start at 80% vac and slowly reduce to 90% or whatever your target vac is. But never from high vac to low vac. I know this is done but the science says its wrong. If you look up the universal gas law you will get some idea of how this works. Pressure is related to volume in this situation so a vol change means a pressure change or a pressure change means a volume change. Ideally you want the volume of your job to stay the same or get smaller as the job progresses.

The biggest enemy of infusion I've come to realise (apart from poor vacuum) is the spring back of the stack. This is always trying to increase the job volume. If you had a rigid mould such as a plastic injection mould these effects are minimal because the mould is metal and can't change volume. But in a soft bag high vacuum condition any micro leak or outgassing gets worse as the bag can get bigger and its a downward spiral thing. I now train people to think of the vac guage as a volume guage rather than a pressure guage. The pressure is important for the process eg styrene vapour pressure and removal of air etc but its more useful to think about volume change and how to stop it. 3) white appearance of CSM seems to be something to do with the binder. I've done multiaxial stacks and CSM in same bag ie in same conditions and CSM is milky. I have two thoughts a) binder needs more resin to dissolve or because it is not aggitated allowing absorbtion and dispersal during infusion it simply sits there hence milky appearance. b) binder is source of moisture which clouds the resin c) binder creates "champagne effect" and creates micro bubbles at its surface creating outgassing or syrene. I have not done CSM with epoxy which would remove this from the list if it worked.

For your first attempts you've done good! From my observations I think the next step in VIP is to use semipermeables (eg goretex or tyvek) as scavenger vacuum edges and part barriers to improve the process. I've done tests with tyvek/breather across parts and the results were very good.

Cheers Peter S

 

Hi Peter, weve now done 4 pieces including the original test panel leading up to this large swim platform, all have included some csm. I stripped it and removed it from the mold yesterday and am very happy with it, its not quite perfect yet but were zeroing in on it. We did not have the milky appearance on the bag side this time I think,as you say, because we did not finish with a mat, the bulk of the laminate of this part is uni. On the last 2 parts I started and finished with a .75oz mat as something to sand when prepping for paint but when you pull the peel ply with mat it tends to lift the fibers a little that makes it look a little more milky than it actually is, the uni on the platform peels clear. We are still getting some small areas that look a little dry. On this part we applied full vacuum, a bit under 29hg at our altitude, to the dry stack for about an hour and then reduced it to about 75% before opening the feed lines and left it there until it exothermed. We did do a drop test and even though we fixed the few leaks we did fine we still were marginal, so we are going to bag the table with some shade cloth and see if we are getting some micro leaks through the table itself. I have a question regarding vacuum pumps, we are using a new 6cfm 2 stage Robinair pump and it spits out a lot of oil from the exhaust, my old pump is a 1cfm Edwards and it always did the same, is this normal?

Steve.

 

Hi Steve W - Yes the HVAC type pumps without oil recovery do spit out a lot of oil. Keep the level up and use a good quality oil. This is due to the water vapour passing through it from the job. When used to pull an aircon down there is relatively little water vapour in the sytem and it rarely has leaks to deal with due to the rigid system and the good connections. In VIP conditions the stack will always have a lot of water in it and due to leaks will pull air through the job continually bringing in water to the job. I put a rag over the outlet or add a tube to a box with a rag etc. I do like this type of pump because you can see the water vapour leaving it. I call this steaming which initially is a good sign. The steaming should end however in 30-40mins so then you know you have removed most of the water from the system. If you can get an absolute vac guage you will find this very helpful. Get one that can run continuously (thermister types run for around 3mins then need to stop for a while, pirahna sensor types can run continuous) If you place this at the resin port side of the system you will see that the pressure pumps down and at about 2000Pa it will stop. This also agrees with the steaming. Water vapour is 10000x more volume then water so in the steaming phase the pressure stops dropping until the vapour stops being produced. Then it will start pumping down again. You can't see this on an analogue guage as it can't measure accurately or low enough. If you look up a saturated pressure chart for water you will also see that to produce steam at various temperatures. So you can look at the temp at the time say 20degs C and the table will say 1650Pa (about from memory) so you need to be under 1650Pa to vapourise the water. The next problem is transport. As you have a bag at say 1000Pa (99.9% vacuum) and the part volume is say 100 litres you still have 1 litre of air in there. People say I have 100% vac but they in fact don't. Most boat builders VIP at less than then 99.9% vacuum (say 5000Pa absolute) so have more air in there then they think. This air however is not a probolem if you have a good layout and a smooth slow fill as the resin front pushes this along and the pump cleans it up as the fill progresses (called vacuum scavenging, I note you use a vacuum scavenging runner around the part very good move). If the flow front is ragged and fast (turbulent) the air gets included into the frow front and gets trapped along the way. At end of the fill this expands a bit and becomes general porosity along with the original air in the resin unless you have degassed the resin prior to fill. I like the two stage pumps as they can pull down lower then single stage pumps. Industrial type and aerospace processes run at about 100-150Pa absolute pressure. ie 99.99% vac. The company I'm working for now and the processes I've developed for them has a specification of 500Pa before a fill can start. These are always measured at the furtherest distance from the vacuum (generally at the resin port) not on the vacuum side of the job as this will always be a good value. If you have the performance chart for the pump you can also look at this curve and figure its dynamic vacuum capacity. A static vacuum is created when you isolate the job from the pump. It should be the same pressure whether the pump is connected or disconnected. If there is a leak on the dry side of the job this leak does not matter if the pump can keep up with it. If the leak is on the wet side of the job it will introduce air into the part once the fill gets to it. Sometimes on non critical parts we take a punt and wait until we see the leak and fix it when we see it in the job. If the leak is on the dry side no problem. But this type of thing does not happen with experience and or good practice or good vac materials. When I train people to build bags a common question is what is an acceptable leak down rate? My answer is if you went to the supermarket and bought some vac sealed food do you expect it to leak down over the next half hour? or day? So the answer is zero leakdown is acceptable. I once did a demo bag on a board for some naval arcitects and they where so surprised it did not come down while siting on their desk for days. They had not experienced that before with their clients. Boatbuilders who build wet bags don't need to be concerned with leakless bags so its a slight head change and gear up to do VIP from wet bagging. But most adapt very quickly such as yourself. I've been doing it for over 20 years now and attitudes have changed hugely to the process. By the looks of it you will be building very high quality parts every time, very soon. My only thought is to start thinking about edge filling vs runner systems. In an edge fill you have the resin runner on the trimmed part so its mark is removed in the trim. You are using a linear fill strategy. so the flow front has a more or less even length. If you fill towards the centre the flow front length gets shorter as the fill progresses (it flows about twice as far as a linear arrangment). This means the fill is faster as the flow rate required gets smaller. It also means the flow front is more reliable. The trick is to put the vacuum in the right place. This can be estimated by finding the longest length of the part and picking midway on this line. eg if rectangular its the middle of the diagonal line. If its a weird shape then have two or three or a small runner at the estimated close out. Cheers Peter s

 

Think about a 3 sided runner first and vac along one side of say a rectangular part. vs having parallel runners on the part. cheers