Is there a general/standard rule to apply to belt/pulley alignment? What is too much etc?
Example: 20 hp motor running at 1750 rpm, running a blower at 1650 rpm. What would be the min/max parallel and angular mis alignment?
Are laser alignment tools the way to go? Is a straight edge good enough?


Thanks

Gates gives the following:
"As a general rule, sheave misaligment on V-belt drives should be less than 1/2 degree or 1/10" per foot of drive center distance.
Misalignment for synchronous, Polyflex and Micro-V belts should be within 1/4 degree or 1/16" per foot of drive center distance."

Also, Sheave face runout should be < 0.005" up to 5" sheave, and 1 mil per inch diameter for larger sheaves. I don't think rim or groove runout can easily be measured.

The above toerance specs listed above are way out in left field. Those tolerances are far too loose; you could eye-ball them.

Run-outs: 0.002" or total stack tolerance of 0.003". On long spans between bearings and where the sheave is close to the bearing you will want 7 mils or less in the center.

Fan/blower: set level. Thermal considerations? Rotor weight sufficient to demand level when in operation? Yes; an offset allowance should be made. If driver has no thermal offset ene-to-end then, consideration must be made for final alignment - level and aligned.

V-belts are matched and as one can see, the width they take-up may span 6" or more dependent on number of belts and size, etc... So to keep all in load sufficiently, alignment must be close.

The angle in two planes should be within 1/60th of one degree or 1 minute (3 mils per 10"). Parallel should be within 5 mils in both planes.

Need more; please contact me. Work day has started.

I have quoted the source for the angle alignment spec. It is Gates - one of the largest suppliers of belts. Page 3 here:
http://www.gates.com/file_display_common.cfm?thispath=G...er%20Alignment%2Epdf

Also, note the tolerance for vee belts is not as tight as for sync belts and it would be overkill to spend the same effort aligning Vee belts as sync belts.

Thank you all for your replies. I must have missed the info on the Gates website as I looked there prior to posting. May I ask what measuring devices do you guys use for aligning belts? Straight edge? Laser? etc..

Thanks Again,

X

A straight edge with 4 points of contact is as good as anything. A string is the least desirable. A laser is good for final tightening and a one man show.
If you haven't used a laser much, offset misalignment will confuse and screw you. After some experiance with it (and learning from mistakes) you can use it from the get go.

Dave

Typical HVAC with undersized commercial shafting gets cranked off-center with set screw or eccentric collar (shudder) lock bearings. Throw in some shaft bend, especially if keyways are cut midspan, and it is possible to get some impressive 1X radial runout at the sheave. The typical 2D fan structure offers little stiffness to resist the 1X tug on the belt, so a highly directional 1X fan mechanical motion and vibration will result. Add an adjustable sheave to the motor for a dash of 1X motor tug, and the fan bearing vibration measurements will create work orders for balance and alignment.

The balance tech will go nuts when his balance weights can only reduce the "vibration" inline with the belts at the expense of the perpendicular measurement. Diligent efforts to improve sheave and belt alignment will not reduce the eccentric sheave cranking.

A dial indicator would reveal the shaft runout and location, and the bearing lock point can sometimes be re-positioned in an effort offset the influence of the undersized shaft. The eccentric adjustable sheave can be a lost cause. Plus their groove pitch/spacing guarantees most of the belts will have a fair amount of offset misalignment.
http://www.adclassix.com/images/57armstrongtires.jpg

quote:

Gates gives the following:
"As a general rule, sheave misaligment on V-belt drives should be less than 1/2 degree or 1/10" per foot of drive center distance.
Misalignment for synchronous, Polyflex and Micro-V belts should be within 1/4 degree or 1/16" per foot of drive center distance."

100 mils per foot can easily be seen by anyone same as 1/2 degree or 30 minutes angle - no need for alignment tools as long as you can see fairly good.

Basically the same can be said for the 1/4 degree as it is a huge amount of misalignment.

For machines with flexible couplings, the coupling manufacturers recommend very loose alignment tolerances, but we all know very well that we should align the machines much tighter than those for long life of the bearings/machines (vs the couplings). The alignment tool folks (such as Ludeca) are understandably pretty vocal in pointing out those valid benefits of tighter tolerances for coupled machines and publish recommended alignment tolerances.

So, does a similar situation exist for belts as for flex couplings (manufacturer's tolerances too loose) ? I am open to the possibility, but I haven't heard anyone suggest tighter alignment tolerances for belts than recommended by the belt manufacturers. The Ludeca laser belt alignment device literature points out recommendations for shaft and sheave runout but is eerily silent on angle alignment tolerances. The Monarch laser alignment tool only mentions the Gates tolerance (0.5 degrees / 0.25 degrees). See the middle of page 4 here
http://www.easylaser.com/GL80%20Instructions%20Sheave%2...with%20photos_2_.pdf

Is there anywhere that tigher angle alignment tolerances for belts are recommended?

Also I am curious if anyone attempts radial runout on the sheave? Is it done on the rim or the groove? We only check shaft radial runout and sheave face runout, but it seems difficult to get any type of radial runout on the sheave itself.

I don't think we look at belts closely enough. there is stretch but elastic properties are on a curve with tension. This make a D belt sheave where 6 belts are employed demand close alignment both parallel and angle amounts.

back w/people walking in - later.

Pete, correct me if I'm wrong, but I think that what Gates is trying to say is that for a general rule of thumb:

V-belt drive angular misalignment should be:

1/2 - degree angle = 1/10-inch per foot of distance traveled....not 1/10-inch per foot of drive center distance

Synchronous belt drive angular misalignment should be:

1/4 - degree angle = 1/16-inch per foot of distance traveled....not 1/16-inch per foot of drive center distance

The example they give is a synchronous belt drive that has a straightedge held flat against one pulley while a measurement is taken from the outside edge of the other pulley back to the straightedge....(see L2 dimension in the attached Word Document)...for a 12" pulley, this measurement should be less than 1/16"....

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

Pulley.doc (42 Kb, 24 downloads) Pulley diagram

Good point Billy. There are some subtleties in the way that half degree misalignment number is treated

The words including "center distance" that you are pointing out from my post 20 July 2008 05:20 PM were taken directly from page 5 (pdf page 7/51) of the attached Gates publication.14995 10/01 (not copyrighted).

Going back to the other Gates document linked 21 July 2008 06:48 AM, on page 3/4 are 2 figures. The top figure shows angle misalignment and is expressed as a ratio of two distances, where the denominator distance is a sheave diameter as suggested in your figure. The bottom figure shows offset misalignment, expressed again as an angle (which at first sounds a little weird, but makes sense when we remember that more offset distance is tolerable across a larger center/center distance). The offset misalignment angle is again the ratio of two distances, where this time the denominator distance is the sheave center-to-center distance. It goes on to say that we add these two angles together (angle misalignment plus offset misalignment) and compare to the 1/2 degree ("When determining if a V-type drive system is aligned within these recommendations, the angular and parallel misalignment must both be measured and quantified individually, and then added together. The total sum of angular and parallel misalignment can then be compared to the belt manufacturer's recommendations for the particular type of drive.")

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

GatesBeltPM_www_gates_com1.pdf (2,429 Kb, 18 downloads)

The idea of adding those two angles (angle and offset misalignment) together doesn't seem to have much physical significance. I suspect it is just important to keep both angles low, and by adding them together before comparing to the limit, we ensure they are low.

Another data point:

"Shaft Alignment Handbook", Third Edition, ISBN 10: 1-57444-721-1 by John Piotrowski states:

He has a little different suggestion for measuring. He does 2 different straightedge comparisons accross 2 different portions of the sheave to detect possible tilt in the other plane . He also allows for the fact that the straightedge may contact each sheave at only one point (so we end up measuring a gap at each sheave).

His "limit" is even higher than the number mentioned above:

quote:

Piotrowski
Typically, sheave misalignment should be less than 1.5 degrees (i.e., 1/8 in. per foot or approximately 10 mils/in.).

Exactly which angle he intends us to apply that to isn't clear to me. I would guess he intends us to add the angle between the straightedge and each sheave.

http://www.coastclassics.com/1964corvairengine.jpg

http://www.uniquecarsandparts.com.au/images/car_info/la..._corvette_engine.jpg

Good point. There certainly are many machines with built-in sheave misalignment.

Piotrowski
Typically, sheave misalignment should be less than 1.5 degrees (i.e., 1/8 in. per foot or approximately 10 mils/in.).



Exactly which angle he intends us to apply that to isn't clear to me. I would guess he intends us to add the angle between the straightedge and each sheave.[/quote]

1.5 degrees is way out in left field. No alignment needed sounds as what that is saying. Totally absurd!

Basically one should level the fan shaft. Ensure the bearing to shaft and bearing to bearing alignment is good. All run-outs must be good, perfect or within tolerance. Now align the motor to the fan - hopefully 3 mils parallel and within 1/60th of one degree or 1 minute. Now consider thermal offset! So, when it gets complicated just ignore everything and open your tolerances - nah!!!!!!! Why not deal with the situation and correct it? I can do it and at the tolerances mentioned in my specification --- not willie nillie garbage. Just plot the _ _ mn thing. Sorry, I'm getting short but I've seen these tolerances come from above mentioned and others that tells me its an excuse for not being able to do it righ. Belts and coupling, shaft-to-shaft as well -- way too loose and not considering many factors.

quote:

There certainly are many machines with built-in sheave misalignment

Those sheaves on the Corvair engine are "aligned" but in 3D, so the belt enters the sheave gently.

Unlike a typical fan The sheave mounting points are also rugged, and interconnected so radial runout and belt thickness variations create "internal" forces but that are within the "endurance" load limits of the bearings.

Many modern cars have rubber bushed alternators to provide some flexibility.

Dan points out something considered by engineering; how to deal with inheirent misalignment.

A car's driveshaft. Car sits on flexible tires and mounted in suspension in a controled enviroment but with limits. One can take that movement and come up with a setting that keeps the driveshaft's angle on the same side of zero all the time thus not allowing slack to be taken up and avoiding breaking the U-joint from torque. But many will take the example from the car and apply it to our industrial setting of fixed machines wrongly claiming gross misalignment must exist so the drive shaft will work properly - nah!!!

Considerations are also made during the design and engineering process for huge thermal growths in our industry. Seemingly, not by all.

Our belt example: claiming 0.5 degree to be OK - under what circumstance. Two fixed machines - nah, should happen. At 1/2 degree there is no need for alignment tools; you can easily see such a gross amount.

Anyone willing to spend a little time with a pencil and graph paper can make sense of aligning two shafts that sit parallel to one another. Typically we set the fixed machine's shaft level. Bearing to shaft and bearing to bearing alignment is performed while the rotor is sitting in its housing with proper seal clearance and back-plate clearance and/or shroud clearance. Angular and parallel is accomplished by sitting the shafts parallel in two planes and the sheaves in alignment with one another. It is not as easy as it may first appear. But one can clearly see getting the fan set correctly really helps the whole process.
Careful inspection must be made of the sheaves even if new. Sheave gages are available for old sheaves not only appearance should be checked. Used the gage to check grove depth as well. Castings are not always equal, so the outside edge of the sheaves may have excessive runout and different thicknesses.

If you are doing a new installation and mounting the motor on an adjustable or tensioning rail, here's a tip: set the motor's shaft parallel to the fan shaft and cant the rail so that the motor's anchor bolts are against the slot on one side and slack is taken up in the motor bolthole as well. Doing this will enable to maintain parallel through the tensioning process and make future belt replacement a snap.

As Dave said, one may opt for straight edge or laser - with care both will do a good job. There is parallel alignment in two planes and any deviation from that provides angular misalignment. So we must have proper angular and parallel alignment with sheaves aligning. I think most of us find that monitoring angular and parallel during the alignment process is easiest and most tools are geared for that type logic.

With proper techniques and procedures achieving 1/60th of one degree and 5 mils parallel isn't so difficult and can be done quickly making it economical. Why settle for eyeball tolerances?

Not really an area that I am working these days. With that said, I shall do what is too often done in such forums - SPEAK OUT.

One thing that seems important to me is the alignment over the sheave diameter, because this is where the belt tension would act (to me - somewhere at the V or cog connections). The belt tension and the angle would seem to place a moment across the sheave, which would affect the shaft and bearing.

I never have calculated the reactions. Are they important? Anyone here do this, or is there something much more important, which I have overlooked? There may be a couple of ways this could put a moment on the sheave.

Many will have noticed that generally during manufacturing of the sheave the groves are turned and the rim is also turned during this process. However, the face may not be 'faced off' and the casting left rough thus making it unsuitable to take measurements from.

During the alignment process; making runout checks, inspecting sheaves (on old units), etc. One can use an outside mic to measure to the top of the gage. Many have a gage to measure belt tension and some may have a gauge for measuring and monitoring. The outside and inside should basically be the same or all should be the same tension.

It would be nice if the OEM or one such as Gates would make additional info available such as Bill described. On my end there is little incentive to do those test as I am looking to good smooth operation and long life and know that by aligning to tight tolerances within economic boundries I'll achieve desired results.

Bill is totally correct in saying the concern is where the rubber meets the road. Our concern is contact regardless if it is a V, micro-V or gear-toothed belt or power grip as some call it.

In our day and age with MCC's and soft starts and step start motors we can find addition applications for the power grip belt. As a rule it last much longer and also assist in extended MTBF with less stress on the bearings. V belts require initial setup and then re-setting in ~2 weeks making the alignment job twice the effort unless you utilize the tip mentioned above.