Last night I ended up staying up a lot later than intended. It’s a long story, so I’ll spare you the details (other than that it involved a cold, a cat, an upcoming open-house, a gassy baby, a LLC competitor, and the RS-68–don’t ask). While I was up, I got reading some of Ross Tierney’s latests posts on Nasaspaceflight.com about his DIRECT concept (he’s only recently reappeared on the scene after a couple month’s absence). One of Ross’s comments got me thinking again.
One of the biggest pieces of evidence used by Dr Stanley in his Q&A thread to call into question the merit of DIRECT was the large discrepancy between the RS-68 Regen numbers Ross had used, and some official numbers given by P&W. Ross had been claiming a vaccum Isp of ~435s or so, and the P&W guys released a formal performance summary stating that they only felt ~418s was reasonably attainable (compared to the existing 409s vacuum Isp). Ross mentioned that while the engineers working with him on the original study had investigated almost 100 different configurations, many of which used bog-standard existing RS-68s, they had taken the gamble of selecting the RS-68R version because they had it from a pretty good source that the higher performance numbers were doable.
One thing Ross mentioned suddenly gelled several different things I had been reading and thinking about over the past several months. Back when I did my quick “Random Thought” post about inflatable nozzles, I had noticed that the original J-2 engine had a vacuum Isp only a little bit above the RS-68, but the new J-2X NASA is working on has almost the same Isp as the SSME (~420s for J-2 and ~448s for J-2X), in spite of it being, I think, a gas-generator driven cycle. The big difference? A much larger expansion ratio. On the simplest level, as you increase the expansion ratio of a nozzle, more of the jet thermal energy is converted into kinetic energy.
The RS-68 and the RS-68R shown in the link above both use very small expansion ratios, while some other ground-lit LOX/LH2 engines use much, much higher expansion ratios. The reason? The RS-68 was designed initially for the Delta-IV. The simplest version of the Delta-IV relies 100% on the thrust of the RS-68 for liftoff. The SSME and Vulcain 2 on the other hand are used on vehicles where large solid rocket boosters provide the majority of the liftoff thrust. The main reason they’re even lit on the ground is to avoid needing to air-start the engine, so all they have to do on the ground is light reliably, and not rip themselves apart due to flow separation. What this all means is that the RS-68 had to be optimized for a much lower altitude than the Vulcain 2 or SSME. This means a lower expansion ratio, and hence a much lower vacuum Isp.
The Ariane-5’s Vulcain engine is a particularly enlightening example. The Vulcain engine is much lower thrust than an RS-68 (~240klbf for Vulcain vs ~700klbf for RS-68), but runs on the same propellants, has a chamber pressure that is very similar (102bar for Vulcain vs ~95 bar for RS-68), runs on a similar mixture ratio (6.2 for Vulcain, 6.0 for RS-68), the Vulcain has a regen cooled chamber with a dump-cooled nozzle, and both are gas generator cycles. However the Vulcain has an area ratio over twice the RS-68’s, so it’s vacuum Isp is 431s vs 409s for the RS-68.
Now, with a concept like DIRECT, most of the liftoff thrust would likely be provided by the SRBs anyway, unlike on the Delta-IV. So it would seem to make a lot of sense to pursue such a nozzle extension. You really don’t need that much extra thrust on liftoff, and the Isp later on would be a lot more useful. Now, I’m not sure if you could package a longer-nozzle version of an RS-68 into the volume required for use with DIRECT, but the numbers look pretty convincing that you could get most of Ross’s “magic Isp numbers” with a completely realistic redesign of the nozzle. You don’t even need to go regen cooled to get most of the benefit. You just need to redesign the nozzle section for a slightly higher expansion ratio. Unless the throat somehow unchokes, downstream changes to a nozzle do not have any real effect on the upstream combustion processes. No new pump work or injectors, or combustion chamber changes would need to be made, just the nozzle skirt. As it is, the RS-68 is already throttleable down to 60%, so they know that they can operate the thing at much reduced exit pressures without flow separation. Doing a nozzle redesign like that for an ablative engine is a much easier task than resurrecting a 40 year-old design like the J-2X. Heck, such a change might even be useful for the core stage on a Delta-IVH.
Now, with all that said, Ross and his group are now focusing on a v2.0 of their DIRECT concept that uses existing RS-68s, SRBs, unstretched shuttle external tanks, etc. So, none of this musing really effects what they’re doing. But I thought it was worth mentioning.
Oh, and I also ended up learning a lot more about nozzle design cleverness. While it’s cool knowing that I’ve been part of a 4-person team that has designed, built, tested, and is about to start flying a fairly darned high-performance throttleable biprop engine, it’s also fun learning more about ways to make those systems even better. Looks like I’ve got a new area to study out once I get this pesky thesis done…
speaking of which…