Post by finiteparts on Nov 3, 2022 20:02:53 GMT -5
Well that was easy...it may not be perfect, but it gives us a gut check...
So here is the 4288 lbf load that I calculated from assuming a 10 inch o.d. and 4 inch i.d. on the backplate subjected to an internal gage pressure of 65 psi...which assumes all of the pressure load is taken through the NGVs, and not the outer casing, etc. total overkill, but it is quick and easy and bounds the issue. I plotted these as only displacement in the x-direction, which can be seen to point to the compressor side. Not seeing the deformation needed to cause your rub, but it did reduce the deflection by about half...more than I thought. (I found a better method to transfer the pictures between computers....sorry for the shitty pictures in the last post)
So let's crank it up to a aft load of 10,000 lbf...still not coming forward enough in the turbine backface area to cause the rub, but getting close. Additionally, the aft NGV band might stiffen things up and cause a different deflection slope near the turbine backface...but my guess is that it would just break the welds first.
So I stand by my assertion that the pressure loads are not you primary failure mechanism.
To be perfectly clear, this is an elastic analysis taken way into where it would have gone plastic, but I think it gives a good sense of the fundamental behaviors of a solid NGV backplate subjected to a strong thermal gradient. Also, the deformation figures are plotted with a scale factor of 19-ish, so in case you're not familiar with FEA plots, these are exaggerated to help make the strain effects easier to see...it is not a realistic image.
Post by finiteparts on Nov 3, 2022 20:58:20 GMT -5
Andy, I snagged an image from one of your videos....Hope you don't mind.
So here it the proposed mod. Cut it along the red line and machine a new center piece that is a snug fit. The purpose of the bulk of that structure is to hold the aft end of the bearing housing centered...there is no need to try to "hold" the axial because you have the bearing housing to do that. The inner portion could be modified to have a cooled cavity to reduce thermal loads into the bearing housing and thus oil, plus it allows axial growth of the housing without inducing an aft load.
I admit that there are going to be some details on the fitment, structural stiffness and local thermal growth, but this seems like a nice method to decouple the thermals of the two parts. You could even include a rabbet fit on it so that it clamps it axially but "free" to grow radially with thermals.
I have about a third of my Tertiary holes being fed by air that has travelled between shaft tunnel and FT inner wall , theres bleed air holes at the "corner" that feeds air to behind the turb wheel in the recess where the NGV securing screws are located , this does an excellent job of keeping things cool in that part as I never have any "colour" flowing back to the shaft .
Once the air turns the corner and flows radially past the NGV vanes there are slight gaps between vanes and the wall slots that the vanes are mounted into , so some bleed air going through there whilst the bulk tries to cool the NGV front wall , the air then discharges through the Tertiary holes as well as slots in the locating sheeting band imgbb.com/58MP5Tk , the slotted locating section of the inner wall is going to run with considerable clearance allowing bleed air supply which should provide some blanket shielding on the other side .
Post by slittlewing on Nov 7, 2022 14:48:53 GMT -5
Thanks a lot for taking the time to put that through your FEA program Chris, you got similar results to mine (not sure about deflection magnitude but at least in shape and direction). And thanks John for confirming a few differences between your latest engine and the kart engine!
I will give it some thought as to how to try and modify the current setup.
Thanks for the confirmation John and for checking your TF15 housing. I will probably have a go at extending the scallop!
One other question, does anyone know what Torque I should use for the HX82 Compressor Nut? The Holset service data sheet says 45Nm for a standard HX82, but there is a one variant 4044033 which should be done at 80Nm.
Given the much larger than standard Compressor, and higher P2 of the engine, I am tempted to go for 80Nm. When I removed the comp, I measured 95Nm to undo the nut, but it had been fitted with loctite.
My HX82 is a 4044033 ,it weights 49kg comes from a cummins qsk45 engine
Post by slittlewing on Sept 4, 2023 15:30:10 GMT -5
A very quiet year on the kart front so far, work was crazy for most of 2023.
I recently designed a new ECU circuit board for the kart, this time with separate high side drive boards as I had a few failures in the past and the cost of components is high to replace the entire board when just one mosfet has failed! Lots more functionality on this ECU compared to the jetbike one. Could be a good winter job to solder it all up.
New afterburner is in build after blowing up the last one. Thankfully I wasn’t too upset about it, as I was planning to make another anyway 😂