Ok.. So I've got some issues still.
After putting the engine together with the new combustor and trying to get her started up, i've come across a few things.
She lights off nice and easy with the propane, that proverbial "pop" as the flame front runs back into the combustor and RPMs start to increase. Ramped her up with compressed air to around 15KRPM, feeding more LPG along the way and started in with the kero.
I installed a small needle valve into the line running to the bearing lube system and had this closed off initially. I also installed a 0.025" jet into the fuel line to help dampen the initial burst of JetA into the system.
A bit of resistance to increasing spool coming in with the kero.. Wanted to throw some moderate flames out of the tailpipe, nothing too concerning but laggier than when she was running well back in the day. Had to keep increasing the jet-a flow, backing off a bit, increasing, backing of a bit, etc to get spool up. Finally get her up to ~60KRPM and ~9-10psi pressure and crack open the bearing lube valve a little. Start getting some frequent "pops" out of the tailpipe, appx 5-8 per second and engine wants to slow down a little. Back off of the bearing lube a tad and add a little more Jet-A. I could get her up to sustain without the need of the air tip into the comp wheel but it was still borderline. Trying to add more fuel to ramp her up resulted in flames out the jetpipe and a hung condition.
In all of these cases, the TOT never exceeded ~8%0C but the ramp-up attempts were very short duration - seeing flames out of the tailpipe just bring back haunting memories, lol.
I went through this process 4 times with a good bit of cooldown and injecting WD40 into the bearing lube port periodically over several minutes of time. When the engine was self sustaining, the bearing tunnel temps were around 174-180C.
It just seemed that any time I attempted to add more fuel, even in very small increases, the burn would not stay within the combustion chamber. Fuel burning past the turbine is obviously going to counter turbine operation and thus resulted in a stalled ramp-up.
I called it quits and pulled the engine down to inspect everything.
This is what I am seeing:
The inner liner is showing good signs of heat all the way up to the discharge of the vaporizer tubes which is good to see. There are no holes in the upper half of the inner liner as I wanted to ensure the combustion was not being blown away from the evap tubes. It looks like that approach is working well.
On the outer liner, it doesn't appear that there is much heat above the top row of primary holes. The metal is slightly darker than the base color (as noted by the furthest forward wall of the combustor - the radiused "head") which tells me it is seeing some heat. Most likely the outward swirl in the primary zone is being blown inward by the top row of primary holes. To me, this is an expected result. The amount of heat the inner liner has been exposed to and how forward this exposure is tells me the primary zone is working as it should.
About 1/3 of the way down the length of the combustor's outer liner, just below the third row of primary holes, the liner is seeing a good bit of heat. If you look through the back end of the combustor, square to the angle of the inner liner's ~30-degree back-wall, through the large 1/2" tertiary dilution holes, you are looking directly at this "hot" zone on the outer liner. Additionally, the secondary zone holes on the inner liner are also at this same axial position as the hot zone seen in the outer liner. All of this makes sense too, although I think the large part of the combustion being blown out towards the outer liner is a result of the latter condition - those secondary holes in the inner liner.
However, directly outward, radially, from the inner 30-degree wall's 1/2" dilution ports, on the outer wall, you see the most amount of heat exposure.
This last observation, combined with the fact that any attempt to ramp up the spool resulted in strong flames out of the jetpipe, I have the suspicion that it is possible that some of the fuel being injected into the vaporizer tubes is actually being sucked back and blown out through the adjacent dilution ports on that same 30-degree inner back-wall of the combustor.
Following the airflow path, the air coming down parallel to the bearing housing, as soon as it comes to that 30-degree back-wall, it has to make an immediate 180-degree turn to enter into the vaporizer tubes. With the large 1/2" dilution ports shouldering each vaporizer tube inlet, I suspect that some of the fuel is being wicked into the flow going out of the dilution ports.
It is apparent that I've got plenty of the combustion being blown onto the vaporizer tubes - much more than I think was actually being dumped onto the v.3 combustor I built for this engine back in 2008. But I think not all of the fuel is actually going into the vaporizers and some is being carried out of the tertiary ports and burning at the NGV inlet and through the turbine, thus resulting in a hung condition.
Additionally, I am speculative of what the distribution of flow mass is between the inner and outer liner with respect to the flow area that the holes offer for the various zones. Just because the flow area is "X" does not mean it will flow "x" amount of air mass.... The actual flow mass value is dependent on the delivery to that port.
I think what may also be exacerbating the problem is that in this new chamber design, I have tapered the inner and outer liner walls - tapered from smaller diameter to larger for the outer liner and from larger to smaller for the inner liner. An attempt to try and equalize the flow distribution into the ports for a 1:1 area:flow relationship.
I think what may be happening is that the flow of air coming out of the diffuser is primarily axial so a larger component of the air mass is going to want to continue traveling back along the outer liner wall and less likely to make the 90-degree turn inward towards the bearing housing, another 90-degree turn axially along the bearing housing, and so on...
A biased condition that limits massflow into the inner volute of the engine is also going to exacerbate the chances of part of the fuel being wicked backwards out of the vaporizer tubes and out through those large dilution ports.
Just FYI, the fuel injector tubes' discharge is situated rather shallow into the vaporizer tubes. I did this to try and use as much available surface area of the vaporizer tubes as possible - and putting a discharge angle into the injectors to try and get a "swirl" effect on the Jet-A to maximize dwell time.
With that logic, I could take two different approaches:
1) lengthen the injector tubes. This will work to minimize the effect of the dynamic flow through the dilution ports to try and suck part of the fuel backwards.
2) Fabricate restrictor plates for the dilution ports that shift the hole center outwards, radially, further away from the vaporizer tube inlets, and reducing their flow area by some fraction, perhaps by 1/2, and then adding the dilution port area back into the outer wall liner.
Additionally, I think the need to install two additional components is warranted:
1) Flow straighteners along the combustor's outer wall
2) Flow distributor to channel airflow towards the inner volute of the combustor.
Adding this last set of elements to the equation adds a degree of re-evaluation to the previous solution matrix though.
Rather than modifying the dilution port flow area on the inner 30-degree back-wall, if flow distributors are added to equalize flow to both the inner and outer liners, it is a good possibility that no modification to the dilution ports is needed. Just add in the flow distributors and the fuel delivery issue is resolved. The flow straighteners are an additional bonus... (you can see in the combustor pictures that there is about a 45-55 degree swirl in the flow, which we all know doesn't work FOR you)...
What I'm coming to realize is that while the hard-and-fast 30/20/50 inducer area ROT is spot-on, one must also consider how the flow through the combustion section will affect the air masses available at each port stage from front to rear as well as how this affects the distribution between the inner and outer liner. The only hard-and-fast rule on this is that all of the other soft-and-slow parameters are where the devil resides!
So, looks like I'll test this theoretical approach by simply adding in flow distributors and flow straighteners and pack her all back together for a testrun!
YAY!
Here's the video:
