Gear software showing negative approach length/ percent?

[RANorrisPE]

Hi All,

I've been looking everywhere I can for this subject without any luck (Dudley, Khiralla, and Buckingham's books and all over the web) without any success.

I'm designing a three gear system, and I'm attempting to make it with long addendums/ all (or at least close to it) recess action since I want to make it as quiet and efficient as possible at higher speeds and it can't be back driven. I'm using geartrax to verify my homebrewed spreadsheet calcs and it keeps telling me my gears have over 100% recess (in the range of 101-116% depending on profile modification). So the recess is longer than the entire contact path. My own calculations come short when I modify the addendum length and I typically use and trust geartrax's numbers for profile modified gears(after I verify this, I intend to work that in), so I attempted to verify by doing a layout of the circles and line of action in solidworks, but I was struggling to verify this. All of the numbers seemed to line up, except the start of action and end of action points.

So my question being, am I seeing an error in geartrax, or is this possible? If so, does this mean the gears won't work, or that they will run even better?

I purchased another of Buckingham's works to try to get more info on this, but I'm not certain if that'll help any more.

My current system is:
Driver: 61T
Idler: 50T
Driven Gear: 40T
Module: 1.125
PA: 20°

I haven't finalized my profile modifications, but my initial point is x1=x2=+1 (1.125mm) and x3=0. The weirdness comes in for this case in between the idler and the driven where Geartrax spits out the following:
Length of Contact: 5.337mm
Approach Length: -0.015mm (-0.29%)
Recess Length: 5.353mm (100.29%)
Contact Ratio: 1.6071

If I add any type of tip relief or tip breaks then things get even further over 100%.

Thanks everyone for your insight!

Ryan

 

[tbuelna]

Have you tried adjusting the center distance? I believe in your case the change in center distance is something like half the sum of the addendum modifications.

 

[spigor]

I took a look at your numbers and it seems the approach should be 29% and recess 71%. Looks like an error in the calcs. There's also a bunch of other problems with center distance and possible interference.

 

[RANorrisPE]

Thanks tbuelna and spigor, I'll check out each of your suggestions.

Spigor, I'm not quite certain how you got those numbers. Can you walk me through what you used/did to get there? I'll run through it again. Geartrax automatically updates the center distance when the profile modification is changed, unless you specify an overridden CD modification. I intend to do center distance extension by a currently undecided amount for backlash, so maybe things will balance out better then.

 

[RANorrisPE]

Just to add - the calculated new/nominal center distance for the gears in question is 51.673mm according to Geartrax. Before profile modification the center distance is 50.625mm. That would be an increase of about 1.048mm (half the sum of modification would end up being (1.125mm+0)/2 = .5625mm, so I think there is something else at play here as well) The modified center distance and the modified pressure angles that Geartrax spits out checks out when I check the layout graphically by modifying the addendum circles.

Thanks again for your insights!

Ryan

 

[spigor]

I got the numbers by means of calculating and I saw, that it should be 29% approach and 71% recess. You have 0.29%, so it looks like the software has not converted decimals to percent, then added decimals to percent and that's a bug in the software. Also, looking at the data I had an impression, that the contact path has been calculated for nominal center distance. If there's + addendum modification only it can lead to interference problems. Since the original question was only about the corectness of software calculation, I looked no further since I believe it's been answered.

 

[RANorrisPE]

Spigor,

Thanks for your reply. Did you do manual/hand calculations or did you use software? I do not tend to put trust in any numbers just presented without understanding where they came from (hence both my initial question, and trying to follow up with yours). I verified that geartrax gave correct contact paths, center distance and approach/recess with no profile shift, and that there was the expected CD change. I'll run through it graphically and work through the vector loops analytically again to try to verify the numbers, and I'll contact camnetics as well.

Thanks again,

Ryan

 

[spigor]

If you want me to prove my numbers, then it's too much work.

 

[tbuelna]

Ryan,

One other question. Are the gear OD's enlarged from standard?