AGMA 9 vs. AGMA 10 6

[hondashadow1100vt]

I am a PE working on a construction contract wherein a number of progressing cavity type positive displacement pumps are being furnished. Our spec requires AGMA 10 rated gearing. The proposed pump manufacturer is acceptable in every way except that they have a standard gear drive offering that utilizes AGMA 9 rated gears. I have been advised that the AGMA 9 gearing is merely built to slightly less precise tolerances than AGMA 10. The purpose of this post is to request feedback regarding what the implications of accepting the AGMA 9 in lieu of the AGMA 10 might be. Thank you in advance for your interest, knowledge and experience!

 

[geesamand]

Yes, this is a reasonable description.

I'm not experienced in pumps, but in general gear reducer service a slightly lower AGMA quality is a non-issue if it's the lower speed gearset and the overall noise and vibration criteria will be observed.

 

[tbuelna]

I agree with geesamand's comments. The AGMA quality number is a minimum standard for dimensional accuracy of the gear tooth active profile surfaces. The main difference between AGMA 9 and AGMA 10 quality gears would be in their noise/vibration and dynamic load characteristics. As geesamand noted, this small difference in precision is not usually a major concern with gearsets having modest PLVs.

You should also remember that the AGMA quality number is a minimum standard for dimensional accuracy, and it is quite likely that a large percentage (>50%) of the gears used in your vendor's pump drives would meet the slightly higher AGMA 10 quality standards. So you might consider asking your pump vendor how much extra they would charge to assemble a set of pump drives for you using gears selected from their production inventory that meet AGMA 10 quality. Since 100% of their production gears meet AGMA 9 standards, it should be no problem for them to find a few that meet the slightly tighter AGMA 10 requirements.

Modern gear manufacturing is one of those areas where technology has greatly improved the quality and precision of the typical finished product. The quality of a gear is largely the result of the machinery used to produce it, and modern CNC gear machines are incredibly precise. So even if you ask for a batch of AGMA 9 gears, you will likely get gears that are much better quality thanks to the improved capabilities of modern gear machinery.

Hope that helps.
Terry

 

[hondashadow1100vt]

Thank you both for your input/feedback on this matter. The following is more detail gleaned from a couple of follow-up discussions.

The manufacturer of the gear box has indicated that the overall gearbox has on the order of half a dozen gears, all except two of which are rated as AGMA 10. The AGMA gears are helical type gears. The AGMA 9 rated gears are two bevel type gears.

In the project's application a right angle gear box is required due to space limitations. This right angle gearbox arrangement is not a typical arrangement for progressing cavity pumps. It is anticipated that the client's AGMA 10 specification is, in part, based on the fact that for helically geared pumps that the AGMA 10 rating is relatively straightforward to obtain with the major gearbox manufacturers.

The gearbox manufacturer has indicated that they cannot provide an off the shelf right angle gear box with all AGMA 10 components due to the fact that it would deviate from their standard manufacturing practices. They have also indicated that they are not willing to hand pick the AGMA 9 rated gear boxes that are also incidentally in compliance with the AGMA 10 tolerances. The gearbox manufacturer has furthermore indicated that for their AGMA 9 bevel gears that they match the two gears together during the manufacturing process and apply some form of compound to the gears to aid that matching process (though I am not really sure that I have fully understood what that compound and manufacturing process entail).

Lastly the manufacturer has indicated that meeting the AGMA gear rating criteria is not really an effective measure of a quality gearbox, suited to the application. They have presented a case that the service factor is really the best measure of the quality of the unit. The manufacturer is furthermore indicating that the bearing life is a better indicator of the anticipated longevity of the drive. The manufacturer is proposing service factors on the order of 2.0 to 3.0 and B-10 bearing life of 100,000-hours.

I think that the questions that I have at this time are:
1. In order to demonstrate equivalency of the manufacturer's proposed drive to the drive that contains AGMA 10 gearing (and they do exist but with a longer lead time), what technical information should the manufacturer present in order to demonstrate this in a clear and obvious manner?
2. Is there any expectation of a shorter gearbox life based on the AGMA 9 gearing?

Thank you very much for you insights and feedback.

 

[geesamand]

" So you might consider asking your pump vendor how much extra they would charge to assemble a set of pump drives for you using gears selected from their production inventory that meet AGMA 10 quality. Since 100% of their production gears meet AGMA 9 standards, it should be no problem for them to find a few that meet the slightly tighter AGMA 10 requirements."

You're assuming they have 100% inspection data on hand for each piece. I would not expect that, at least from the industries where I have experience.

I'm not sure I agree with everything the gear mfr is saying.
- "bearing life is better indicator of drive life" L10 bearing life calculations are wrong. Well, you should aware that they can be wildly inaccurate depending on lubrication and speed factors that the old/simple L10 calculation ignores. (L10a gives you a more complete understanding of the life of the bearings, but that's not necessarily a trustworthy statistic from the manufacturer). I guess my overall point is that published bearing life that goes beyond the warranty period is not worth much.
- On the same point, if the gear design calculations are not conservative, the gear teeth can wear out (pitting) long before the bearings spall. Bearing life would be no indication of life in this case.
- Higher service factors can help, but a manufacturer who is steering you away from <2.0 SF in a shock-free application gives me the impression their gear ratings are junky and need padding. Not that you should buy a 1.0 SF, but if a 1.5 or 2.0 is not enough then their ratings could be, well, aggressive.
- AGMA rating is certainly not the only factor in drive quality. It is strictly a control over the accuracy of the tooth forms. It does not control bearing quality, gearing design, housing accuracy, housing rigidity, shaft rigidity, or any of the other critical factors in gearbox construction. It's actually quite easy to design or build AGMA Q15 gears to run like Q7s.
- Bevel gear quality is a tough situation for manufacturers. They are describing a cut/carburized/lapped manufacturing process, which tops out around AGMA 9 for Gleason cut or AGMA 10 for Klingelnberg cut teeth. After lapping, the gear and pinion take on a shape that matches each other, and make much better contact and run more smoothly, but technically do not increase in measurable quality. But lapping works and this Q9 manufacturing process is very effective for the low speed gear reduction.
To go up in bevel gear quality the next option is Klingelnberg hard-cut or Gleason ground bevel sets. These are very expensive, usually long-lead time in small runs, and come in at AGMA 12-13 quality. I suspect that would be a waste of your resources.
So my point is that you can't buy bevels in Q10 or Q11 for a bit more than Q9. Your options are basically Q9 at a reasonable price or Q12 for a small mint.
If another manufacturer also has bevel gearing but claims Q10 for them, it's probably a lie or their sales team isn't actually aware of the realities of bevel set quality.
For many years customers have requested Q10 gearing as a standard practice, and for many years they have been receiving right-angle gearboxes with Q10 helical high-speed and intermediate reductions and Q9 bevel low speed set with no ill effect or damaged expectations. Some manufacturers are open and honest about this technicality, and others pass over this white lie. My company makes this kind of right-angle design for a non-pump application and I will attest that it's a sound design. Just don't put that Q9 bevel set on the highest speed reduction.
- I don't think I would push the mfr to create a special manufacturing process for you to get nominally "better" component quality. There are so many aspects to gear reducer quality that if a manufacturing process is working well, it's best to leave it alone. Specials are in reality a perturbation to the manufacturing process and can lead to other quality or lead time problems.

It sounds like you have a couple of manufacturers making basically the same thing. They may or may not be an equal comparison. I would not punish the one who is being open about their bevel gear quality. Or if the other mfr is claiming a quality advantage, ask them to provide their inspection reports of their Q10 bevel gear sets.

I think if you want to control quality, you should focus on 1) mfr reputation 2) mfr quality control (not the stated gear quality, but their control of the process) 3) stated noise and vibration limits 4) warranty 5) tooth contact patterns under load

David

 

[tbuelna]

Just to be clear:

A) this is a 3-stage right angle reduction gearbox, where the first two stages are external helical gears, and the third stage is some form of helical bevel gear mesh like a spiral bevel.
B) the first two (high speed) external helical gear stages use AGMA 10 quality gears
C) the third (low speed) spiral bevel gear stage uses AGMA 9 quality gears that are adjusted at installation to provide an optimized contact pattern.

If this is the case, then I think you should be OK. As noted previously, the minor difference between AGMA 9 and AGMA 10 quality gears would only be an issue with noise/vibration or with dynamic loads when high PLVs are involved. Since the AGMA 9 gears are only used in the low speed stage, and these gears are carefully rigged/shimmed at installation to compensate for some of the manufacturing errors, I don't think you should expect a reduced L10 system life based on this fact alone.

Showing a multi-stage gearbox being good by analysis for an L10 system life of 100K hours is no simple matter. It would require very modest stress levels in the components, low load cycle frequencies, and/or very high quality materials used for the components. A difference between AGMA 9 and AGMA 10 quality in the low speed gear stage would have negligible effect on the system L10 life compared to other variables like service factor, lubrication conditions or material quality.

 

[hondashadow1100vt]

David (Geesamand), Terry (tbuelna):

Thank you both very much for the insightful feedback. This is one of those times when I think that I ought to go back to school.

I gather that I have a little homework to do here. I will dig up the manufacturer's gear construction information (e.g., a cut away of the gears may do the trick) to verify the gear arrangement (and speeds) in order to confirm that the gear arrangement is in fact a three stage right angle gear box where the high speed gearing is in fact handled by the AGMA 10 gearing. I do recall that the bevel is spiral type and that this is the gearing that only meets the AGMA 9.

I am greatly comforted by the primer regarding the basics of AGMA 9 Gleason cut versus AGMA 10 Kingelnberg cut teeth. It really helpful to have some qualitative insights regarding the cost and relative quality expectations regarding what is being proposed versus what is being asked for by the spec.

I also agreed that as a matter of general practice I find it is typically best to go with the manufacturers standard practice (and spec that from the beginning) rather than try to force their hand into altering their normal way of doing things, except when the manufacturer specifically offers the special feature as an option on their pre-printed standard literature. I am glad that your commentary have reinforced that practice in this specific area of engineering/construction.

Thank you very much.

Sincerely,
Mark

 

[tbuelna]

Mark-

No need to go back to school just to get a better understanding of the technical issues involved with your gearbox design. Plenty of resources available to learn about them on your own.

If you want to quantify the impact of AGMA quality number on a gear mesh, one way would be to compare relative AGMA Dynamic Factors (Kv) for the Q10 and Q9 gearsets under identical load & speed operating conditions. As previously noted, gear quality becomes more important as PLV increases. This is due to increased dynamic loading caused by transmission errors at the mesh contact. Attached is an (older) chart showing Kv adjustment factors for various levels of transmission error (Qv) in the gear mesh versus PLV. If you want to learn more about how to calculate Kv for your gears, then get a current copy of AGMA 2001-D04. The analytical approach is given in section 8.3.

Good luck to you.
Terry

 

[hondashadow1100vt]

tbudelna:
Thank you for the chart.
I will seek out access to AGMA 2001-D04, section 8.3.
Thank you again,
-Mark

 

[verknde]

I would like to add the following point under consideration:

While the gear quality of cylindrical involute gears are based on the measurement of profile, lead, pitch and run-out, the bevel gear quality is only defined by the measurement of the pitch (single, cumulative and pitch error) and run-out. This is valid for AGMA, ISO, DIN, JIS etc. The most important character of bevel gears, which means the contact pattern (and topography), are not defined and not classified by any kind of norm.
The contact pattern (size and position without load and expectation with load) have been designed by the manufacturer and his experiences for the calculation of the final layout of the bevel gear set.
The measured bevel gear quality can be fantastic, but bad contact pattern influences the smooth run of the gearbox, noise, wear etc.

By the way, new Klingelnberg bevel gear Generators (latest machine generation) are able to finish hardened bevel gear sets (diameter > 1000 mm) in practice in DIN 3/4 that means approx. AGMA 13/14.
With regard to Gleason machines (generators and grinders) I have too little current experiences.
5x milled bevel gear sets > 1000 mm outer diameter achieve constantly DIN 5, AGMA 11 and better for smaller diameters (depending on the tool wear during the finishing cycles).

 

[tbuelna]

verknde-

You make an excellent point about the importance of contact pattern quality of spiral bevel gears under operating speed/load conditions. With high-performance spiral bevel gears it is often not even adequate to check the contact pattern with a slow roll-thru under load. Since the lube film conditions present at operating speeds will produce a different result than a slow roll-thru does. Designing and manufacturing high-performance spiral bevel gear sets has always involved a bit of trial and error. The availability of modern gear design software tools has made the process easier, but it still is a job best left to those with experience.