What a very good question. I have a better one. Have you got the gear tooth form and information that would allow one to describe a construct of the helix? The problem I have had in the past is getting that information. I know that there are available data on involute gear tooth forms based on SAE standards etc, but mostly there are costs associated with accessing this information.
Once supplied and understood, then the short answer is that it is quite probably a swept section of some sort with an emphasis on controlling the section orientation relative to the helix and quite possibly the central axis of you gear.
Helices are supported in NX so it should not be too difficult to describe the geometry. I don't know enough about helical gears to be able to address whether the section is taken normal to the axis of gear rotation or swept normal along the helix.
I took a model for a straight cut gear (reasonable facsimile) and used Insert>Sweep>Swept to add a helical twist using a straight line (gear axis) and Angular Law>Linear, 0-20 degrees, to create a single tooth then arrayed to create the full gear.
This is a very simplified example based on a low speed, low load, low precision gear (by a non-gearhead). As Hudson said, for true geometry you will have to rely on existing standards. The sketch that describes the tooth profile may need to be oriented normal to the helix to get a true section shape.
that comes from a gear manufacturer's website and is credited as such being available as such placed in what in my limited understanding should be deemed public availability such that I should be able to share it with you.
Michael,
I'm not certain of your involute tooth form or whether it ought really to be aligned with the direction of the sweeping rather than on the plane of the side of the gear. Some of the gear material that I have read seems to indicate that but I'm yet to manage to check it. Since the attached is based on what I believe to be ISO module 3 gear form so if I can find that information which certainly available somewhere then we ought be able to check and find out according to what this manufacturer has done.
What is clear from your model is that although the method is quite correct 20 degrees of rotation about the central axis does not equate to a 20 degree angle. I think that the amount of rotation is to do with an chord length that derives from the thickness of the gear and the given angle in this case 20 degrees. Then you have to rotate that according to a calculation that uses the radius of the gear to rotate the chord length distance. The trick is where to strike the chord length whether it is at the OD, the ID or some contact point in between the two?
I have spent some short time looking for information about this without much luck. Should badpeanut have that to hand then it may be of some use to us.
There is also a grip program that models gear forms that has been found on this site, but is has a large number of input fields that I don't have the information to use properly or understand.
This gives you the shape of a gear as a DXF out of Excel would you believe it!
I had another method but it felt the results contradicted the manufacturer's file that I'd obtained and found this closer, so it may help.
There's a couple of versions of a gear modeling grip program here.
I got these to compile, but don't know how to use them yet.
If anyone ever comes up any instructions about the inputs that you need to make this grip program work I'd like to know. I roughly follow how the ISO metric gear tooth modules work, but the values that you need for this other program aren't immediately familiar to me. Surely you'd only want to put in a couple of simple numbers to arrive at a sensible result? There are a dozen or so here which seems more than you would hope to have to input.
I also compiled those two grip programs but they create non-parametric models and they are not for helical gears.
I will check out your Excel to DXF link.
I think we are getting close to having a parametric helical gear using a law curve to generate the involute tooth profile, and a helix to sweep the gear cross section over.
I was giving this some more thought and came up with this.
What is a helical gear but an infinite number of 0 thickness straight gears slightly rotated in relation to each other? If that is the case, then the tooth profile would indeed be perpendicular to the central axis of the gear and swept in the manner of the model submitted.
True the tooth profile in the model submitted is not a true involute, just close enough for its intended use. As for the 20 degree angle of rotation, I pulled that number out of thin air. The actual rotation angle would be a function of the thickness of the gear, the helix angle and the pitch diameter.
No the grip programs aren't for helical gears but I may have forgotten to mention that the tooth profile does appear to work according to the 2D form on the side of the gear, not any such similar profile on a plane normal to the helix. Because of that the profile that is developed by the grip programs is still good for something. As I was working through this problem for my own interest I wondered if I could figure out how the GRIP worked by comparison with the MIT excel program so that I could confirm my understanding.
Michael,
I agree with your construction method. It appears the same as the manufacturer's model if I plug in the parameters matching to their construction. What I'm still trying to fathom is how the correct figures were calculated. So far I have sourced three profiles for the same basic gear criteria Module 0.3 with 20 teeth and they're all just slightly different.
