Gates belt theroy questions

After using the WCP belt calculator and learning the math behind it from the Gates belt theroy book I found on discrepancy that caught my attention. The calculations were for a HTD 5 180 tooth timing belt between two 24 tooth pulleys.

The math that I ended up with using the math in the Gates theroy book was:
pitch*(Belt teeth - pulley teeth)/2 = c to c distance
5mm*(180t-24t)/2 = 390mm

The under lying equation from WCP seemed to be:
pitch*(Belt teeth - pulley teeth)/2 +0.001"= c to c distance

So my question is what is the point of that extra .001" added to the distance?

Sent from my phone(quickly during class…)

Pretension. Make a tighter belt. Prevent belt from flying off?
Most FRC teams can’t even manufacture to the tolerance of 0.001" anyways…

Generally, additions like that (or to the C-C distances on gearboxes) are made to account for manufacturing tolerances. By fudging it a tiny bit large, you give yourself room for error while still getting a tight belt. (If it was too tight, you could always fudge the hole back a little and survive. It’s far harder to make that happen when it’s too loose.) That particular number would come from the calculator maker’s past experiences, and may vary for other applications.

I’d say the number of teams that can make that happen grows annually (and maybe beyond the FRC growth rate) as precision manufacturing gets cheaper and more widespread. Whether that ability is in-house and/or whether they take full advantage of it is another topic.

See page 182 in the Gates Power Drive Design manual. The standard pre-tensioning center distance should be in the .02 to .03 inch range for most HTD or GT3 belting sizes used in robotics.

Hell no, that is significantly overtensioned. If you take a closer look at that document you’ll see that they aren’t referring to pretensioning distances at all:

Since fixed center drives are not recommended, center distance allowances for a Gates PowerGrip® belt drive are necessary to assure that the belt can be installed with-out damage and then tensioned correctly. The standard installation allowance is the minimum decrease in center distance required to install a belt when flanged sprockets are removed from their shafts for belt installation. This is shown in the first column of Table 4. This table also lists the min-imum increase in center distance required to assure that a belt can be properly tensioned over its normal lifetime.

The table immediately below that you got the 0.02" and 0.03" numbers from refers to the minimum amount adjustability (think versablock) in c-c distances that they recommend so that the belt is easy to install and so that it can be tensioned as it stretches over time (as timing belts often do in industrial applications).

In FRC, however, fixed c-c distances work great because of the limited run-time of our robots, and are generally designed to be exact or a couple of thousandths large.

The Nominal Center Distance is for very light loads, it is a zero belt tension condition. Also consider that belt lengths have tolerances (typically +/- 0.08 inches for short belts). With a Drive Train with two CIMs near stall torque; you will skip teeth if you don’t have adequate tension. The only options are a tensioner or extra length in the Center Distance to insure a preload.

The belts are normally rated for 8000 hours, so run time isn’t a FRC issue. Adequate Tension for the required load and speeds is a concern for reliability.

Can you give a reference on this? I don’t see anything on the VEX website. If I were to buy a belt on SDP-SI, they give tolerances in the range of 0.02 in. Which belts or suppliers are you talking about?

That’s the Gates specificifation. It would still apply to GT3 and their HTD. The HTD patent is expired, so there’s lots of aftermarket HTD belts. How more or less accurate second source suppliers are is unknown. But for as cheap as these belts are now, I wouldn’t expect tighter tolerances.

Their GT3 Belt Drive Design Manual lists C-C belt length tolerances on the order of 0.01 in

I believe 2791 just runs direct C-C with no center distance addition going off what Chris is Me says. Although the tolerance may be 0.02", realistically it could be much tighter than that.

Not super scientific, but we generally run .005" over on our loaded belt runs. Even that is in the range where it starts to get hard to install in an exact C-C application.

Also… Theroy Jenkins!

Yup, we don’t add anything to the center distance and it “just works”.

With regards to the specs - the center distance equations are the same. You’ll notice the WCP calculator has a field called “center distance adder” that is set to .001 by default - that is where the term comes from. Leaving this at one thou is fine if you want. Doesn’t seem to make much difference either way. If you needed to be within one thou of a specific exact distance for belts to work, you wouldn’t be able to run them with exact centers anyway because almost nobody holds those tolerances in FRC.

As for the rest of the specs - keep in mind they are referring to mechanisms with far higher runtimes than FRC things. They are very conservative.

Thank you for all of the information. Would the pure calculation or adding a few thousandth to the distance yeild a drive that would be less likely to snap a belt?

Also I have heard that mixing different belt with htd5 pulleys can result in less failure than just using HTD 5. Is there any merit to that?

Once you add enough distance, you start to over-tension the belt, and you risk failure that way. Once you remove enough distance, you risk ratcheting.

5 thou or less added seems like it would be fine.

Using GT2 belts with HTD pulleys is slightly stronger than HTD / HTD, but weaker than GT2 / GT2. I don’t think you need to do this - if your belt strength is marginal enough for this to make the difference you don’t have a big enough safety factor.

Remember, 15mm wide for 24T pulleys in a drivetrain.

This brochure from Gates does a pretty good job of explaining what belts interface with what pulleys. The one thing it doesn’t include is the GT3 series of timing belts, which interface with GT2 pulleys.