I am at a customer site and need to find out if this shaft is running anywhere near critical.
SHaft is 2.937" Dia., 35.94" total length. It is an overhung configuration.
Bearings are at 11" and 24.5" in from sheave end of shaft.
Wheel is 32.69 inches diameter. Center is at 33.5".
Weight of wheel is 150"

Does anyone have some software you can plug these into and come up with a critical speed. They are running it now at 3000 to 3600 RPM.

Dave

This message has been edited. Last edited by: RRS_Dave,

A wheel weight is essential. Then you can plug this into a canned formula from Machinery Handbook as a minimum and get a get "guess-estimate."

John,

I was in such a hurry I forgot that. It is 150#.
I'll add it up above.
Thnaks

Dave

Dave,

Did you figure this out. If not, I'll dig my Handbook out of the box. Let me know.

Danny

No Danny,

If you have a handbook, whip it on me (formula) and I'll calc.

Dave

Dave,

I e-mailed you scans of the pages in the handbook.
Gotta take the old ball and chain out to dinner.

Good Luck,

Danny

Dave,

Here is a link where you can download different programs in EXCEL that will calculate critical speed, etc. of rotors.
You get 1 month of free usage then if you find it useful, you can buy the 'key' and use it forever.
The link is the HOME page. You'll have to go down the page and search for SHAFT. First, you'll have to register yourself before downloading. A very fine sets of various mechanical calculation sheets.

Good luck,

MarkoLeo

http://www.mitcalc.com/en/download.htm

Here are my results using an old version of critspd transfer matrix program (better double-check the results with another source in case I made an error)

Let Kb = stiffness of each bearing including support stiffness.

Kb=5E6 lbf/inch -> F1 = 9040rpm
Kb=0.5E6 lbf/inch -> F1 = 5250rpm
Kb=0.1E6 lbf/inch -> F1 = 2640rpm

API684 lists rolling bearing stiffness as 5E6 but this seems very high and does not include any support stiffness.

My guess is typical rolling bearing plus support is in neighborhood of 0.5E6 but just a guess. Others here would know better.

Maybe you can describe the bearing support?

RRSD.ppt (80 Kb, 23 downloads)

Attached is some info from the last run which shows the model in more detail including options selected. The only thing I changed between the three runs was stiffness of both bearings.

I calculated disk thickness as 0.63" by assuming it was a solid steel disk with diameter 32.69" weighing 150 pounds.

RRSD1.txt (3 Kb, 15 downloads)

Thank all of you for your replies,

Danny, those are the same formula I have in my old (35 year old) statics and machine design books. I'll take you up on doubling your money. I don't have a machinists handbook. I always use the one over at the shop. Will you take a credit card?

Markeolo,
Thank you for the link. I have downloaded and will try it for a while. This looks like the analysis given to me from an engineer at the plant I was at. I am sure he was using the same thing. However, He did not have the fan on there, only the shaft. I am currently hunting for the place to put in the rotating mass of the fan. Thanks again.

EPete,

Thee are the typical two bearings on an overhung load. They are SKF bearings, and they are not rated for this speed. As happens in a lot of places, this fan was designed for a max speed of 2400 CPM. However, they have changed some production variables downstream, and reversed the sheaves, and now run it at somewhere in the 3000 to 3600 RPM. It is variable speed, and once they go above 3000, phase shifts 90 to 180 (bounces back and forth), and it shakes like mad. The whole fan is sitting on four spring isolators, but after doing a ODS on the system, it appears only two corners are actually being isolated.
The bearings sit on top of a triangular type base, made of stainless and it appears about 1/4" thick. I have attached a VERY quick and dirty drawing. In teh ODS, they move around at 5 mils, and appear to be doing a circular dance.
I have trouble trying to figure the stiffness of the bearing, and don't really know how to. I suppose I will have to do some homework!
Again, thanks for all of your help. It is appreciated.

Dave

This message has been edited. Last edited by: RRS_Dave,

GB17_Pressure_fan_drawing.doc (26 Kb, 24 downloads)

Another resonance to check would be the mounting resonance for the springs. It may be possible that the bounce mode is isolated, and a rocking mode is not isolated.

Can you test this by impacting? Also, check the relative phases (and mode shape and/or operating deflection shape) at the corners. Vertical transducers may work, but you should check 3-D.

An overhung pivotal mode would like to interact with a rocking mode, and a pivotal mode might drop in frequency with a foundation rocking mode.

Bill,

The ODS resulted in an apparent rocking between the motor and fan housing with phase being out 180 degrees. It appeared the middle of the frame was flexing. We see belt freq's on this unit, and have checked alignment several times. The apparent rocking between them would put tension on the belts and slack and I believe this is where the belt freqs come from.
The motor appears to be doing more rocking back and forth than the fan pedastle. It appears the motor is loose from the take up base, but danged if we can find it. I suspect flex at the slots where the bolts tie the motor to the frame. We can't see any cracks, but didn't take the motor up from the base either.
I have playing with the software Markeolo turned me onto, and keep coming up with a critical freq. of around 2000 RPM, with the second critical being right around the first. I can't find what I am doing wrong, but will keep fooling with it.
Thanks for your response. I appreciate your input. We will be doing bump tests next shutdown.

Dave

Just from a rough look, it seems like if the machine were somewhat rigidly mounted, the critical speed the effective bearing/support stiffness would be 0.5 E6 or more and the critical speed would be well above 3600. It looks like maybe the the spring isolators (if working right) could lower the bearing/support stiffness enough to lower the critical speed at or below 3600 rpm. I can repeat the calcs with a lower range of stiffnesses to see how it looks (although I'm headed out and won't post anymore until tonight). For best results should also add the sheave mass - do you know how much it weighs or approx dimensions?

One idea that pops into my mind is trying to run with the isolators intentionally blocked to make them as rigid as possible. I'm not sure if there is any easy way to do it or if it would be safe.

Here is an article on somewhat similar overhung fan on rolling bearings written by some members of the forum - Ted Myrick and Barry Crawford. I think they had a rigid base and stiffness around 0.2 E6 lbf/inch. Lots of good information.

TurboTAMU_OverhungFanCriticalSpeedT33pg033.pdf (191 Kb, 27 downloads)

Sheave is 7.1", I am going to guess around 18 to 20#.
I think it is a 5V, 4 belt sheave, probably 5-6"wide.

This software I am using really shows the difference in criticals when varying the stiffness coefficient "k" for the bearings.

I downloaded the article. Some of your and Bills previous posts on forward and backward whirl are beginning to make a little more sense now about what it is. And I am glad to find others have found different resonances when spinning the wheel. We have suspected clearance issues when we found that in the past, but never had (took is a better word) the time to take the wheel off and investigate.
Thanks for your interest.

Dave

This message has been edited. Last edited by: RRS_Dave,

EPete,

The software I downloaded (MITCALC) uses Rayleighs formula for determining shaft criticals. How does the "k" in that formula relate to the Kb you are using? Are they basically the same??

They have a drop down menu selection to pick the type of bearings with type of load, and when I picked freely rolling bearings with overhung load, the k value is 0.9.
I found the lookup table they were using and changed that to 0.5 and the resonant frequency dropped from 5,040 down to 3020 I think (I'm on my laptop now in a motel, not at home where I did the analysis).

How do the two "K's" relate? Or do they?

Thanks again for the schooling.

Dave