Huge, Humbling Hurtle

Vacuum Bagging Cowl Parts

---

Note: You wouldn’t know it by subscribing to this blog, but I have actually been continually working on things in the Garaggio. My frustrations over the failures I was experiencing with vacuum bagging and how demotivating that was made it difficult enough to keep working on projects much less write volumes of blogs over failed attempts. Mixed in over this time, we had a condition inspection on Betty, a few workshop organization projects, finishing my welding table, and small bits of future Legacy projects. So rest assured I am still very actively working on the Legacy, Betty, and mostly Sid (Brian’s Defiant).

Why vacuum bag if it has been such a hurtle?

The light weight, lofted, high quality, molded carbon airframe of Brian’s Defiant was a major reason I wanted to become its caretaker. Those traits made me excited about the project and inspired to continue in the same vein. The vacuum bagging process was integral to achieving the quality and light weight results that became a hallmark of Brian’s work.

• 

• 

• 

Putting a part in a bag and sucking the air out of it seems like it should be straight forward and relatively simple. For me, it has been a huge, humbling hurtle.

Being strong willed (Mom doesn’t like me to say stubborn because it always begets the question, ‘from whom I got that trait?’ and lest I get in trouble, remember ‘whom’ is both singular and plural), I wasn’t willing to give up and just stick with wet layup parts. I figure, Brian’s Defiant should have better.

Vacuum bagging is a skill I want in my repertoire. It is a superior way to make composite parts. I have spent a significant amount of time on youtube, consulting with experts who have been generous with their time, looking over catalogues and technical data sheets, and reading all I can about vacuum bagging. With all this, I consider myself to have more than average technical competentence with the theory. But significantly lacking in practical experience.

If you didn’t buy any of that, I started down the vacuum bagging path with this project and it kicked my butt. There is no way my strong-willed self is going to let vacuum bagging get the better of me.

Just how many failures does it take?

I’m reminded of TV commercials from my childhood with an animated owl asking,

“How many licks does it take to get to the center of a Tootsie Roll Pop? 1, 2, 10. The world will never know. “

It probably would have been better for me not to know. But with my (im)patience, I was counting. I had 13 failures before I had what I would call success. I won’t list them all off, but the picture below shows 4 of the eventual 9 identical induction inlets I made. All of which were failures. Some of them are useable parts, with enough finishing work. But a vacuum-bagged, molded part should have a finish much, much better than this.

What was the cause of the failures?

It took me a while to figure it out. While each had their own nuance, the root cause of all the failures was one thing, poor quality molds. My molds were virtually incapable of sealing without further work. Even if I did a perfect job of applying a leak-free bag to the mold, I was going to have leaks through the mold. Leaks, that in some cases, the vacuum pumps couldn’t overcome. Sure, I also had the occasional poor technique when sealing a bag that caused a massive leak. The majority of leaks I experienced were in the molds themselves.

The aft engine lower cowl mold was made in two pieces and has a seam down the middle, that requires sealing in preparation for a vacuum bag. That one is easy to understand having a massive leak. But all the molds, even my tiny little inlet mold, leaked. Why?

• I made all of these molds in the hottest part of summer. Usually single-handed, I was working against a quickly curing pot of epoxy in 100+ degree temperatures.
• To add insult to poor ideas, I also used a minimal number of plies. Because of how much work and time making a mold requires, I wanted to use only the amount of material that was absolutely required to make what I figured would be an effective, single-use mold.
• Additionally, I didn’t have the proper tools. I should have had bubble busters (serrated rollers) to remove air bubbles from chopped strand mat that was a large percentage of the thickness of the laminate.
• The little inlet mold had epoxy-sealed plywood as the sealing surface. Even with saturating it with epoxy, plywood is porous and a poor choice of material for a mold intended for vacuum bagging.

If you are a composite person, you are grinning at my folly and could have predicted the outcomes. End result is I ended up with molds made of thin laminates, tons of pin holes, and didn’t set myself up for a high probability of success.

How do we avoid or fix these problems?

The answer to that is somewhat simple. Make the best quality molds you can, no matter if they are one-use molds or intended to pull 100 parts. That may not be as actionable of advice as you want, but it is basic and good advice. Don’t do a mold layup in 110 degrees, with epoxy, using minimal mold thickness. Maybe plan to do your mold making in a multiple layup process, or in an environmentally controlled space so you can lower temperature. Whatever you do, try to ensure that you plan a mold that minimizes the possibility of leaks. This could be,

• using tooling gel coat
• making sure you have a thicker mold laminate that minimizes the possibility of pin hole leaks
• having the proper tools so that you can minimize air bubbles in thicker plies such as chopped strand mat
• Using more modern materials such as tooling dough if you are laying up a mold

If you end up with a mold you suspect has leaks, there are ways to possibly fix or work around that. They are band-aids, but they may salvage the mold you made. They did for me.

• You can always add layups to the exterior of the mold. A secondary layup may seal pin holes.
• You can possibly put the entire mold “inside the bag.” This essentially eliminates the requirement for the laminate surface of the mold to be air tight. The challenge for me with this was not only size of the mold, but also that many mold features would likely tear a bag.
• Paint the exterior of the mold with multiple coats of flex-seal. This is what I ended up doing.

My first success.

I have paraphrased quite a bit of the last few months of vacuum bagging trials and tribulations. This has been quite the learning curve for me. But it has been worth it. My first vacuum bagging success was the fiberglass tool cowl from the aft engine upper cowl mold. Far from perfect, but the best I’d done to this point.

The exuberance I felt when we achieved vacuum on this part, literally had me hopping and hollering. I felt validated for not giving up on the process. When I saw that we were able to draw 25″ of vacuum on this part and that I was reasonably certain that we were going to end up with the expected results, I was elated.

I want to give credit to Greg and Sebastian who were integral parts of doing this layup. It took the three of us 3.5 hours to wet out the fiberglass, put on the bagging consumables, and draw the vacuum. We worked out a very comfortable and predictable method and set of roles to make this layup and vacuum bag a success. I was pumped.

The next day we pulled the part from the mold and I was even more excited. We had fantastic surface quality, good laminate compaction, and what was to be expected from a vacuum bagged part.

Seems over-zealous to be so excited about a sacrificial part meant only to be a template or a tool. But it did allow us to fit the cowl to the airplane, trim mating edges, and come up with trim lines using lasers, sharpies, and chalk lines that will represent the aft edge of the cowl. Additionally, we were able to locate an oil check and servicing door as well as cleco holes that will allow us to precisely locate future copies of the cowl to the fuselage with minimal fuss.

• 

• 

• 

• 

• 

• 

• 

• 

• 

Once we had the tool cowl representative of what we wanted the final version to be, we put the cowl back in the mold and used a sharpened scribe to trace the perimeter and cleco holes into the mold. These scribe lines will transfer to subsequent parts and make any copies of this mold repeatable. The oil check door location made a reference for us to apply sheet wax that made a joggle in the mold. This joggle will become the sealing surface and flange for the oil check door which will be a separate part.

Second success.

Greg, Sebastian, and I did a rinse and repeat of the molding and vacuum bagging process. The fiberglass tool cowl was a great dress rehearsal that we just needed to mimic. This layup we used the carbon fiber that we were planning on using for the final part.

If I was elated with the tool cowl, but when we pulled this carbon flying part from the mold I was jumping for joy. Literally. I was literally jumping up and down and pumping my fists. We achieved vacuum. We controlled the pull down of the vacuum to have good consumable placement. We did the layup in good time.

• 

• 

• 

• 

• 

When we pulled the part from the mold, the surface was as expected, not perfect but very good. We had excellent compaction of the layers. We had a good oil servicing door joggle. The scribe lines transferred.The cowl was appropriately stiff. And surprisingly, it was amazingly light. I knew it would be light, but it was surprsing how light it was. I haven’t weighed it yet, but I can’t wait to see where it comes in.

• 

• 

• 

• 

• 

• 

• 

Has it been worth it?

Unequivocally, yes. It certainly is. Now. I may not have said that 6 weeks ago, but I think it has been time well spent. I put the above word success in italics, if you noticed, because I didn’t achieve leak free vacuum and there is improvement that can be made through further learning. But we have been able to achieve at least the 95% result of a perfect vacuum bagged cowl. With my experience level with this method, that is more than I had hoped for.