Shaft Critical Speed 1

[fredt]

Does anyone know of a low cost program for calculating shaft critical speed. Ideally one that takes support stiffness into account as well as being able to input various masses and shaft diameters.

 

[alexit]

Check with University of Aachen, Germany - program was called SPILAD. I have used v2.9 and it was available in English or German versions.

 

[tgmcg]

http://www.turbosynthesis.com

 

[electricpete]

That program is $795. I guess "low cost" is a relative term.

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[bubbaplus]

What is the importance of the half critical speed? I have heard that it is an issue in long slender shafts -- l/d 15 to 25 -- but have not personally seen stabilty problems under that condition. I keep hearing people insist on it as a design parameter, but none has given me a reason or experience to support the concern.

 

[fredt]

The fact that this question is asked scares me a little. Critical speed in rotating machines - especially large rotating machines that can kill people - is as important as a pressure limit is to a large pressure vessel. It ranks equally with structural integrity as a design parameter in large rotating machines.

Operating a machine at it's critical speed in many if not most cases with large machines will result in disintegration within a few seconds. At best, if the machine is heavily damped it may continue to run but with very high vibration levels.

In some classes of small machine the dimensions of the components, driven by other criteria, are relatively so stiff that critical speed is far above running speed and in these cases there may be a tendancy to ignores the issue. However in large high speed rotating machines the shaft dimensions are almost exclusively driven by the critical speed parameter. Generally if this is satisfied shaft stress will be a secondary issue but one that still must be considered.

The accurate calculation of critical speed is quite complex if all factors are taken into account and is possibly the principal component of the engineering speciality known as rotor dynamics.

 

[electricpete]

fredt - What scares you? Do you have some insight into "half-critical speed"?

=====================================
Eng-tips forums: The best place on the web for engineering discussions.

 

[sms]

fredt, all I can say is man ease up a bit. Are you sure you are not confusing over speed with critical speed? There are many turbine generator sets in industry that run right on their critical speed and have for years with no problem. If there is sufficient damping a critical speed is not critical. In addition even if the damping is not great, if it is well balanced, it still is not a problem. The class of machines I speak of are a bit balance sensitive, but it is not a big deal. We run high speed compressors that have enough damping and are so well balanced that the critical speed amplitude is less than 1 mil pp (25 microns), I could run those machines on their critical for hours or days with no problem.

Over speed protection is important, an overspeed will disintigrate a machine and can kill, and overspeed protection could be related to a pressure limit in a vessel, but not a critical speed. Just to make it clear overspeed= running the machine faster than it was designed to run, this will cause it to come apart dangerously. Critical speed= the speed at which a rotor natural frequncy is equal to the rotor turning speed. This can result in high vibration, but if well balance, and well damped, is usually not a problem.

As for the other question, I think it is a bit confused as well. I like to avoid having a critical at half the normal operating speed, since whirl instability comes in at a nominal 0.42 to 0.48X running speed. Whirl is typically manageable, but when the whirl instability excites the first critical, often refered to as whip, then a resonance condition occurs, and the very nature of the whirl instability is to negate the damping. Theoretically if whip occurs then the vibration amplitude should approach infinity, and fredt's concern is then founded. But practically, the rotor contacts the stationary elements and gains damping from that, limiting the vibration, but damaging the machine. If the first crtical is say 60% of normal running speed, there is no way for a whirl instability to excite it, and the problem of whip is avoided. The flip side is that you would like to avoid having a critical (ususally the second) at twice the normal running speed. There are certain machine malfunctions like severe misalignment, and shaft or support stiffness assymetry that can generate a 2X running speed forcing function. If there is a critical at 2X running speed it will then be excited. Again this may not be much of a problem if it is nicely damped, but second criticals rarely are so a little bit of 2X forcing function gets greatly magnified and may pose a problem. As a result, from a design stand point it is good practice to avoid haveing a crtical at twice running speed.

 

[sms]

Oh, and getting back to the original question, you can download an evaluation version of XL rotor at:

http://www.xlrotor.com/index.shtml

I don't know what the program actually costs, but the evaluation version is pretty powerful on its own....

 

[fredt]

I am well aware of the difference between critical speed and overspeed. What I meant by structural integrity in my original response was the ability of the machine to handle the centrifugal stresses.

I stand by my original statement. While I agree that it is theoretically possible to run at critical speed with near perfect balance and adequate damping - it is not a perfect world. Most large simply supported machines cannot be safely run at or very close to critical speed. I can attest to that from personal experience and to believe differently is potentially dangerous.

I have found few problems running near 2X critical speed but none the less agree that it should also be avoided.

 

[fredt]

Just a correction to my previous statement - I should have said few problems running at half critical speed.

 

[bubbaplus]

Let me clarify a few things about critical speed and "half critical speed" and ask for new response. First, I have seen the effects of running at/near critical speed and have done many tests of overspeed trip systems and will do everything I can to avoid either condition.

More specifics about the "half critical speed" situation that I am interested in include:
1. The speed I am inquiring about is 1/2 the 1st critical. That means that if the 1st critical is 1200 rpm, then I am interested in what happens at nominal 600 rpm.
2. The applications I am inquiring about are usually in the operating range of 200 to 600 rpm.
3. Bearings are either spherical roller or deep ball.

I was particularly interested in the whirl phenomen that sms described. I have seen pronounced vibration near half the first critical speed in steam turbines with sleeve bearings with operating speed of 3000 rpm and up and understand that to be "oil whirl". Is this typical in other bearing types and what is are good design parameters relative to 1st critical with other bearing types?

 

[sms]

Well Fred, I guess we will have to agree to disagree. I don't see any great problem with centrifugal stesses at the rotor critical speed, while that is definately the problem at overspeed. The real concern at a lateral critical speed is that if the balance is off, and there is not enough damping, the rotor will contact the stationary elements and do damage, maybe even wreck the machine if let go down that path, but even then centrifugal stresses are not the issue. The issue is pure interference between rotating and stationary parts.

I will also say that excitation of a torsional critical is a very bad problem, torsionals tend to have almost no damping, and can easily result in shaft fracture and complete destruction of the equipment, but then that would be torsional stresses not centrifugal.

Let me just clarify as well, I am not advocating that machinery be allowed as a general rule to run right on a critical. It is certainly good design practice to seperate operating speeds from critical speeds. I also strongly recommend that rotating machinery design be carried out by someone very experienced in rotordynamics. All I am saying is that the state of machine design has advanced to the point that critical speeds are not so critical anymore, and building a rotor model in one of these programs is a great way for a maintenance or operating engineer to get some insight into the behavior of his equipment. But I also work in the power and petrochem business. Perhaps the equipment Fred works with is not so robust.

 

[alexit]

Half-critical speed has to do with mode shape, this causes "whip" as described by Sms. Our shafts run very close to 1st critical (~20% below) all times, we see vibration peak at 1/2 critical but not to exceed our limits ~6mm/s so we permit operation here.

As Sms finishes, there are many vibration modes engineer should be concerned about: centrifugal, linear, torsional each have consequences. The program Spilad permits calculation of all three modes I list. It is more bad for our machines to run at critical cenrifugal than critical linear (assembly fits to rotor.) Our shafts have values like 1st linear = 0.2-0.4x 1st centrifugal = 0.45-0.65 1st torsional. We have not run machines while at torsional critical speed before.

 

[fredt]

Re. sms response. There is no direct connection between critical speed and centrifugal stresses as I am sure that we are aware and I did not state that there was. The comments relate to the issue of overspeed that you introduced and only serves to obscure the issue. The only relevance of rotor (Not specifically shaft) centrifugal stress is that it is of equal importance to the integrity of the machine. I suggest that this issue should be dropped from this particular debate.

I am glad that we agree that running at or near critical speed is not a good thing. My experience relates to large fans with rotors up to and exceeding 20T generally with speeds not exceeding 1000RPM in the case of the larger machines.

I also agree that hydrodynamic bearing oil whirl can be a source of vibration (if the bearings are incorrectly designed) and coupling misalignment is another source among others.

Regarding the issues raised by alexit. Running about 20% below the true critical should be OK. My only reservation is that some variables EG stiffness of a concrete foundation can be difficult to quantify accurately at the design stage and a little more margin is perhaps prudent.

Torsional critical is less of a problem in the machines that I deal with but sometimes is an issue. Usually it is easiest to tune out by varying the coupling stiffeness (and many couplings do have damping) but variable speed drives do complicate the issue both in regard to exciting forces and avoiding a speed band.

 

[sms]

fredt, I'm sorry, I obviously mis read your post of 14 Jul 05 at 9:40.

"I am well aware of the difference between critical speed and overspeed. What I meant by structural integrity in my original response was the ability of the machine to handle the centrifugal stresses."

I thought that refered to centrifugal stresses at the critical speed. It was not my intention to cast aspersions or insult you in any way. If I did so, please accept my apologies.

 

[fredt]

Thanks sms

As you may have gathered critical speed related issues are an area that I feel quite strongly about. I believe that they are a source of many problems - at least on large fans - and not as deterministic as most people think owing to hard to quantify variables.

 

[GregLocock]

Half speed whirl (as I was taught it) can be excited by non axisymmetric shaft stiffness, causing a 2/rev excitation, non axisymmetric rotors, or by oil films. So far as I know true oil film whirl tends to occur at 'around' 2/rev, the first two should be exact. I'd guess oil film whip is only likely in plain bearings, not roller or ball bearings.

I have also heard half speed whirl referred to as double speed whirl.

Cheers

Greg Locock

Please see FAQ731-376 for tips on how to make the best use of Eng-Tips.