Bearing failure rating? Simple rev counter?

[agusta]

Firstly I would like to point out I have absolutely no idea what I am doing and will most likely be blind or dead by tomorrow. But anyway.

I've finally finished me first turbo jet engine (from scratch I would like to add and designed myself and dodgy as I am pretty convinced that it won't explode (mainly because its about a hundred times stronger than it needs to be) but I do expect it to either disintegrate or have bearings fail. I have used roller bearings, which was a stupid idea I know but it saved me a lot of time. What I would like to know is how I can determine roughly what speed a roller bearing will fail at excluding sideways force and heat etc. I have a formula for the failure rating when a certain sized bearing is under a sideways force but I can't find one for the maximum revs they can handle.

There must be a formula where you can use the ball size, number of balls, outside diameter, inside diameter width etc.

Also I would like to have a simple way of working out the number of revs its doing. Maybe using a strobe light and adjusting the frequency until the impeller stops moving? Then you could use that frequency someway. Just a thought. (probably stupid but hay I am at high school where making a fire poker in engineering is classed as complicated)

The thing is I don’t what it disintegrating and blinding a few fellow class mate. I suppose I would get on the news if it exploded and kill a few people (I shouldn’t say that, sorry)

I know I am only at high school and you guys are all probably either 4th year mechanical engineering students or actual engineers but I hope to be a 4th year mechanical engineering student one day so give use a break)

So if you have any ideas that’d be great. Anything really.

Cheers

 

[FredGarvin]

I have used roller bearings, which was a stupid idea I know but it saved me a lot of time. What I would like to know is how I can determine roughly what speed a roller bearing will fail at excluding sideways force and heat etc. I have a formula for the failure rating when a certain sized bearing is under a sideways force but I can't find one for the maximum revs they can handle.

An initial guidline for bearings you can use is a number called the dN number. It is a relative measure of a bearing's speed tolerance. The dN number is the OD of the bearing (in mm) multiplied by the RPM value. This will give you a number that you should be able to compare. Manufacturers, if they don't have it already published, will give you that if you call them. Basically, the more below that number you are, the better off you'll be.

NOTE: Please understand that this is for the speed ONLY! It does not even come close to looking at loads that the bearing will see in actual use. Bearing loads are complex analyses.

Your statement about using a roller bearing is not a very good sign and I'll tell you why. You are going to have, at the minimum, two bearings in your application. A forward and a rear bearing. One of those bearings, usually the aft bearing, can be a cylindrical roller bearing (I'm not going to get into why now). The other HAS to be some kind of angular contact bearing. The reason being is that your compressor (and fan if you had one) are gooing to produc a thrust load on the shaft assembly, i.e. it's going to want to push forward. Your forward bearing is what is going to have to take that load. Plain roller bearings can take some axial loads. Depending on your size, it should be OK. However, in a lot of cases angular contacts are used because of their increased load carrying capacities. I certainly hope you have considered this in your construction. If not, don't stand anywhere near this thing when you run it up.

Also I would like to have a simple way of working out the number of revs its doing. Maybe using a strobe light and adjusting the frequency until the impeller stops moving?

Yes. The strobe is the cheapest way you can go for that. We use accelerometers to monitor the vibrations produced. Part of that signal will be the one per rev signal which is your angular speed. However, I don't think you'll be wanting to do that. Again though, using a strobe means you'll have to be near this thing when it's running and I would NOT recommend that until you have a lot of confidence in this thing.

Then you could use that frequency someway.

Yes. That signal's frequency will most likely be your main component of imbalance of the rotating hardware and its magnitude usually the best indicator of general health of the engine. Higher one per rev vibes indicate that your bearings are having to take a greater radial or longitudinal load and that means they are working much harder than they have to which means decreased life.

The thing is I don’t what it disintegrating and blinding a few fellow class mate. I suppose I would get on the news if it exploded and kill a few people

If this is some kind of project for school, I highly suggest that you take a moment and consider just how much energy is going to be involved in the rotating group of your engine. Coming from someone who does this for a living, even extremely small rotors can and will do trmendous damage if they let loose at speed. You had better take every single precaution possible, INCLUDING not being anywhere near it when you start it (especially stay out of the plane of rotation). All it takes is to see one turbine let loose and go through a cinderblock wall to get the idea of the potential bomb you are dealing with.

Out of curiosity, what exactly did you "design" yourself in this? Did you design the layout? I am assuming you have used a compressor and turbine that are in something like a turbocharger or some already made design?

 

[Danger]

As Fred emphasized, be extremely careful with this thing. If possible, fire it up remotely from behind a massive barricade. (Better yet, have it inside a very strong room or building.) There are optical tachometers available that just watch the rotating part and give a digital read-out, but I don't know how much they cost. Don't even think about standing beside it with a timing light. I remember that when I was a little kid a jet passenger liner (DC10 or something) shelled a turbine on the runway. It cut the bloody plane in half.

 

[FredGarvin]

So at a guess what RPM could a 15mm OD and 6mm ID bearing handle? 5000 10000 15000? Would 10000 be a safe guess? Or is that way too much?

There is no guess for what DN you can expect to be able to run at. There are too many bearing design points that effect that allowable value such as cage material and configuration, roller element material and bearing clearances. The only way to know is to get the part number from the bearing and get ahold of the manufacturer.

I've just built a steel cover over top of it out of a bit of 5mm steel tube cut in half. It was a bit of industrial steam pipe and it looks pretty strong. So now if any thing goes wrong there should hopefully only be two ends for things to fly out of.

There's one safety step that you were right in doing. A scatter shield is always a must.

Would you be able to tell me all the design flaws so that I’ll know why it doesn’t work.

Sure, but I'd need to see what the internals really look like, especially the flowpaths.

I mean is there know way it could work because its so wrong or is it feasible?

There are a lot of reasons why it might not work. The best you can do is try it and troubleshoot as you go. There's no way right now I can tell you if it will or will not work. Give it a try and if it doesn't work AND you walk away from the experience, post what happened and we'll go through it.

I know that the flame will keep burning because I tried it without any blades and using the vacuum cleaner to get the air through.

That is a good idea.

So, the big questions I have for you are:

- Who/how did you balance the rotating group hardware?
- How are you governing the fuel flow?
- Where are you going to run this?

 

[agusta]

So, the big questions I have for you are:
- Who/how did you balance the rotating group hardware?
- How are you governing the fuel flow?
- Where are you going to run this?

I have a friend who is a turbo mechanic so he balanced the shaft with the impellers on it but at the time I hadn’t made the spacer which now runs between the bearings (to keep everything tight but still free, in theory) that may be a problem. I think it should be balanced well enough for what I want.

I’m using LPG and have a pretty good gas regulator that I’ll use. I know LPG is quite flammable and a little dangerous. Thats why I am going to have air running through the engine all the time (using the compressor or vacuum cleaner) and I’ll keep a flame burning at the back all the time as well. Sound like a good or bad idea? Hopefully I won't get a big build up of gas and have it suddenly explode.

It's at school at the moment so I think when I try to get it going I’ll do it at lunch time when no ones around.

I had a bit of trouble making an igniter to light the gas. I had a spark plug and a coil which made a really good spark but would light the gas at all. I also made a glow plug which actually worked really well, when it got red it lit the gas straight away. The thing is it cools down too quickly so if you have air running through it to get it started it cools down and won't ignite the gas. Any ideas for a good igniter?

Cheers

 

[FredGarvin]

Well, it appears that you are doing a pretty good job covering the basics. LPG is probably the safer way to go than what I have heard. Plus you don't have to worry about vaporizing or atomizing the fuel.

As far as the igniter goes, you're going to have to experiment. Have you tried anything akin to a pilot light? If you can use a separate LPG source and regulator, get a small flame going in the burner and hope the airflow doesn't extinguish it. If it does, you still would need a spark source of some kind to constantly be going off until you had a sustained burn. Perhaps an old car distributor set up?

I also forgot to ask, what are you using for a bearing lubrication scheme?

 

[Danger]

Agusta, that's a much more professional-looking thing than I was expecting. Looks like you did a fine job of it. Good luck.