|
Post by enginewhisperer on Feb 4, 2024 23:15:02 GMT -5
these stresses seem much worse than I'd have guessed.
How does it compare to things like car engine flywheels or auto transmission flex plates? They're often ~300mm diameter and ~2mm thick with a starter ring gear on the outside - which is often pressed on! Most of them don't need to handle 10krpm, but they often do 8-9krpm, and practically everything does well over 4000rpm.
|
|
|
Post by Johansson on Feb 5, 2024 0:18:32 GMT -5
I didn't have much time to work on this model over the weekend...I had to fix my Honda Civic that had a charging system issue. The electronic load detector failed and the ECU would not command the alternator to start charging. Of course, no one had the part, so I had to hit up the junkyards to find one and get the car going. I was able to run some rotor speed checks on the main disk and unfortunately because of the limited capability of the steel, the rotational speed that it can support is also limited. S355 has a yield stress of 355 MPa (51 ksi), which is rather low in the rotating disk world (713LC yield is around 116 ksi). I had to bring the rotational speed down to 4500 rpm to just get the 80 mm bore close to yield. But I am not sure that the weld will carry the stress very well, so if I assume the 110 mm bore of the outer disk, then the bore will start to yield at 4200 rpm. The problem with the weld is, I am not sure if you have a full penetration weld between the inner and outer disk. If it is just welded on one side, then the weld can't carry the full load. The plot of the stresses on the outer disk is shown below: You can see at the bore radius ( ~ 2.165 in) the hoop stress exceeds the 51ksi yield stress, and thus the bore will yield out somewhat. This is just the load due solely to the weight of the disk itself, and does not include the stresses that would be imposed due to the pull of the blades and the bolts. I will try to get some more time to work the numbers with the blades and bolts loads. But it will be bad. The bolt holes in the disk will cause local stress concentrations (stress concentration factor (Kt) ~ 2 to 2.5). This is due to the fact that we have removed material, to make the hole, that would have otherwise been carrying some of the hoop load. So , even if it was out at a radius of 5.5 inches where the hoop stress is lower, with a Kt the local stress may be 40 - 50 ksi, right back up to the materials yield stress. If there are any imperfections in the bore or the bolt holes, there will be additional stress concentrations...this is why the bore of a turbine disk is so critical and sensitive to machining finishes, nicks, etc. Again sorry to be the bearer of bad news...I had hoped to get it up in the 5000-6000 rpm range, but that is not likely. I will look to see if there are any other cheap materials that may be more capable. The other thing I saw was that the material capability falls rapidly around 500C and so you would need to keep it cool also. Chris Hi Chris, This is fashinating, I could never have imagined that there would be such loads on a steel disc spinning at "only" 10k rpm. No need to be sorry, even 4.000rpm is well above idle so I will get it running no matter how bad the end numbers are. Just to make sure that you got it right: The steel disc has a bore of 50mm, and the hub is welded onto the side of the disc. Thanks a lot for the help, this is very educational for me. 4.000 rpm is like lawn mover revs, is it difficult to grasp that this solid steel disc would start disintegrating at those speeds. /Anders
|
|
|
Post by enginewhisperer on Feb 5, 2024 0:54:20 GMT -5
VW Beetle cooling fans are a simple sheet and tab / weld construction, and they run up close to 10,000rpm in standard form. People have apparently run them faster than that too. offroadvw.net/tech/wes/fan.html
|
|
|
Post by finiteparts on Feb 5, 2024 8:54:37 GMT -5
FYI....I am only concerned with the point of yield, not burst. This is the point when the bore of the disk will begin to plasticly deform and potentially grow in size. Since you were using what looked like an expansion coupling, if the bore yields out there may be loss of concentricity and torque capability at that point.
I don't think you have to worry about burst. I think the disk/blades will be more likely to stretch out and fowl the casing before they get to a burst speed.
The VW impellers are smaller, ~9 inches in diameter...smaller diameters mean lower rotational speed (radians/sec) and the centrifugal force is a function of the square of the rotational speed....so seeming small radius changes can have a big impact on the forces involved.
As for flywheels....drag racers use billet flywheels. A billet has much higher material stress capabilities. 4140 Chromoly when quenched and tempered has a yield strength of 986 MPa (143 ksi)...much more capable.
- Chris
|
|
richardm
Senior Member
Joined: June 2022
Posts: 336
|
Post by richardm on Feb 5, 2024 10:39:29 GMT -5
I m not too much educated in those complex calculations but I would dare to add the 'imbalance' factor. I guess all this calculations assume a perfectly balance disc assembly and that those forces are evenly applied on the parts of the assembly.Any imbalance will induce even more force on a specific area.. But I might be wrong..
|
|
|
Post by Johansson on Feb 5, 2024 10:57:29 GMT -5
FYI....I am only concerned with the point of yield, not burst. This is the point when the bore of the disk will begin to plasticly deform and potentially grow in size. Since you were using what looked like an expansion coupling, if the bore yields out there may be loss of concentricity and torque capability at that point. I don't think you have to worry about burst. I think the disk/blades will be more likely to stretch out and fowl the casing before they get to a burst speed. The VW impellers are smaller, ~9 inches in diameter...smaller diameters mean lower rotational speed (radians/sec) and the centrifugal force is a function of the square of the rotational speed....so seeming small radius changes can have a big impact on the forces involved. As for flywheels....drag racers use billet flywheels. A billet has much higher material stress capabilities. 4140 Chromoly when quenched and tempered has a yield strength of 986 MPa (143 ksi)...much more capable. - Chris Hi Chris, Ok, I understand. With a tight radial clearance I will have a nice indication that things are starting to stretch beyond what is healthy when the turbine blades start rubbing the casing. Looking at the first page in this thread you can see the von Ohain experimental engine, using a simple riveted sheet metal rotor 610mm in diameter they got 300lbs of thrust at 10.000rpm. I sort of assumed when reading this that the revs weren´t an issue, as long as the bearings could cope with it. Since I have made 90% of the rotor already I will finish it, get it balanced and see how things go. If you can figure out an RPM number when things are about to go south badly I will appreciate it. If a rotor rub is the worst that can happen I will probably try to stretch the limits a bit once everything is up and running, once the new improved rotor is built the ones I have now will just become wall art anyway. /Anders
|
|
|
Post by wannabebuilderuk on Feb 5, 2024 11:01:24 GMT -5
I say finish it as cheaply as you can and get enough cameras/data logging sensors set up and let her rip (slowly at first 😁) and see how/if parts fail and improve on that.
|
|
richardm
Senior Member
Joined: June 2022
Posts: 336
|
Post by richardm on Feb 5, 2024 11:35:44 GMT -5
Quite amazing that the von Ohain experimental engine made it to 10 K given the technology of the time. Could all those calculations be too restrictive?... Did Ohain ignore all those numbers and just gave it a try and succeed? May be you could just go the Ohain way ... about the same way way you see it. Give a try staying safe.
|
|
|
Post by Johansson on Feb 5, 2024 14:54:17 GMT -5
As long as it won´t grenade on me I am no stranger to running engines to failure, the only way to find out where the limits are. Chris: I have been thinking about the mk2 rotor assembly, and one of the main goals with this engine is to prove that it can be built by amateur means. Having an inconel rotor sort of goes against that goal, so what do you think about the idea to figure out a rotor design that is constructed from readily available materials (steel, stainless, high tensile aluminum) and still can cope with pretty decent revs? Perhaps 10k is hoping for too much but maybe 7-8k is within reach? /Anders
|
|
richardm
Senior Member
Joined: June 2022
Posts: 336
|
Post by richardm on Feb 5, 2024 15:20:51 GMT -5
How about a ring welded to the disc over to bolt circle to reinforce that area ?
|
|
|
Post by Johansson on Feb 5, 2024 16:02:08 GMT -5
How about a ring welded to the disc over to bolt circle to reinforce that area ? The problem with that is that you add more weight that tries to rip the hub apart. The same if I go the other way and make the disc thinner to lower the centrifugal load, then there is less material that holds the rotor together. Perhaps the optimal budget design is a set with a solid stainless bladed disc with one high tensile steel disc fitted to each side of it like a sandwich, held together by rivets or bolts. The parts can be laser cut from sheet metal and have a thick hub where the cone fitting sits. /Anders
|
|
ripp
Veteran Member
I'm sorry, I don't speak english, so I torment you (and myself) with a translation program,Sorry
Joined: January 2013
Posts: 236
|
Post by ripp on Feb 6, 2024 5:19:22 GMT -5
Hi Anders,
maybe you should do it like the FD3/64?
if you weld a nut on the outside and a threaded sleeve on the inside The turbine disk does not contain a bore hole.
If you secure with a grub screw after screwing the shaft to the turbine disk, you will have a secure connection.
Ralph
|
|
|
Post by Johansson on Feb 6, 2024 16:46:27 GMT -5
Perhaps, but I like the idea of being able to slide the turbine wheel onto the shaft and clamp it down. Hopefully I will be able to make a sturdy enough turbine wheel without having it boreless.
|
|
|
Post by racket on Feb 6, 2024 18:30:40 GMT -5
Hi Anders
I like your sandwiched idea , the thick discs could be tapered so as theres less mass at the bigger radius but thick at the bore , sorta like the superback on our comp wheels which keep the bores from stretching at high rpm.
Cheers John
|
|
|
Post by finiteparts on Feb 6, 2024 21:42:15 GMT -5
I think you are confusing the HeS 1 engine with his demonstration model that you show. If you read page 18 you can see the demo engine only ran with a electric motor and eventually overheated. If you look at the cross-section of the HeS-1 on page 20, you will see that the rotor 'disks' are more substantial.
I dug up his biography, "Hans von Ohain, Elegance in Flight" by Margret Conner to see if it talked any more about this demo engine. On pg 39, it reads " The engine did not self sustain but did result in the unloading of the electric motor." After this, he had no money left to move the project forward and that is why his Professor put him in contact with Heinkel....which is where the HeS-1 comes to be.
The idea of a sheet metal rotor is sort of still a challenge...at any significant speeds there just isn't enough material at the bore to carry the weight of all the stuff radially outward. But perhaps you could turn a thick plate into a more tapered rotor like shape?
|
|