Their chemists were able to work out an approach to essentially “paint” the fibers with a polymer. You could dial in the density of polymer desired to impregnate the ceramic substrate. “There are two big knobs, the substrate that you’re using and to what density and the other big knob is what polymer that you would impregnate in and to what density,” Rasky said.
He told us that this would give you a fair amount of range in the type of materials that you can produce depending upon the application you have in mind.
Actually, our chemist came up with it very quickly. We had a very good chemist working with the scientists. She is one of the co-inventors, Dr. Min Tae Soo, and when we explained to her what we wanted, that we wanted to put polymer into the ceramic substrate ... and essentially ‘paint’ the fibers, it was only about a week or two, she came up with the way to do it.
That’s when Dr. Soo came up with the idea of diluting the polymer into a solvent and do the impregnation into the substrate, then polymerize the impregnant before you remove the solvent.
I commented to Rasky asking if that once you remove the solvent, there would now be a fine coating of material left, and he answered:
Well, we’ve looked at the topology and the morphology in the material and it turns out you get some interesting structures depending on some of the other additives that are put in it. You can get kind of web structures or you get a variety of structures of the polymer with regard to the fibers. So it’s somewhat complex, but the bottom line is that you have a nice distribution of polymer through the material. And what that does, it turns out, is it has a key aspect or a key feature that makes this work for thermal protection system (TPS) and that permeability drops dramatically from just the ceramic substrate.
In other words, the ceramic substrate themselves, because they’re largely void space, 89 percent void space, they have quite a bit of permeability, air and other things can get in them. And once they go through the impregnation process, their permeability drops dramatically. And that’s important to not allow hot gases to come in when you use them for thermal protection systems.
As compared to what NASA used on the Apollo program as a heat shield, APOLLO ablator AVCOAT 5026, PICA brings down the touch labor cost significantly, Rasky told us.
That one has a lot of hand work because you have a fine honeycomb that you first have to bond carefully to the TPS carrier structure and then you have to fill each cell. And that's one of the things that I know our Orion project has hoped to automate, to do robotically. And that takes a lot of hand-work.
Whereas, you're right, the PICA and SIRCA are both billet fabrication approaches, so more similar to standard hand-automated manufacturing where they use, what they call, foam blocks for interior spacing materials. And so you can make these things in big billets. The PICA, it's only limited by the size of the carbon billet that you can get.
We were buying carbon billets at SpaceX that were roughly 24 inches by 42 inches by about 15 inches and so we can impregnate entire billets. In fact, they have a capability there to do six billets in one batch, so they could develop lots of bulk material quickly and then you cut out the individual pieces that you want out of those billets, like you would out of a foam block and then attach that to the carrier structure. So, from that standpoint, yeah, it reduced the touch labor considerably compared to a hand filled AVCOAT 5026.
XIRCA also has some cost-saving aspects. “There is now a blanket that's similar to what we actually flew on the shuttle," Rasky said. "There was a thing called AFRSI, Advanced Flexible Reusable Surface Insulation, and what SpaceX did is take that basic AFRSI blanket and worked out a way to do a silicone impregnation based off of our patent that we have for XIRCA.
What that allows them to do then is to have these flexible blankets on the aft heat shield of Dragon. It turns out on these space vehicles the aft heat shield, the one that's downstream from the hottest heating, is usually where you have all your closeouts. You have complex seals, and thrusters for your reaction control system. You typically have doors and panels. And they were running into a lot of cost of what are called closeouts.
Maybe you can have some rigid TPS over parts of the door, but where the door connects with other doors or other pieces, you have a gap and that's where you need to do what's called a closeout. It's usually a flexible seal of one type or another that you fill in. Same way those things are expensive and what they've been able to do is go to XIRCA, which is a big flexible blanket that essentially functions as both the acreage cover for the panel and also the closeout. So again it's dropped down their touch labor cost considerably on the aft heat shield. So they're apparently quite happy with both the PICA and the XIRCA.
I thought that XIRCA was a pretty clever modification of SIRCA by SpaceX to exactly fit their need on Dragon. They initially used a material called Acusil II, which came from a company for which Rasky had formerly worked, ITT-Aerotherm. Acusil II is called a syntactic, a foam silicone polymer, that has both silica micro-balloons and fibers stirred into it. It is applied as a kind of paste, Rasky told us, onto a carrier structure and then you vacuum-bag it, cure it, and machine the odd mold lines on it. This was used on the first versions of Dragon.
SpaceX found that it was quite expensive and also quite a schedule driver. They first looked into making some of their own rigid materials, syntactic foams. They do have that now as well, but on top of that they went to the impregnated blankets, XIRCA, and found that this has cut their cost considerably.