TurboCAD Forums

The Ultimate Resource for TurboCAD Knowledge

Register
 
Interested in some really terrific mobile apps? Visit www.turboapps.com for details.

Extruded helical bevel gear
Read 7019 times
* September 30, 2013, 09:35:36 PM
Any idea's as to how to achieve the bevel helical gear in turbocad v19.2 pro plat, I can apply the twist without any problems, but am baffled as to how the draft angle works, it just seems to angle all the sides of the gear profile at that angle rather than the extrusion at that angle. I have never had any success with draft angle with any of the extrusion or sweeping methods, apart from circles or retangles and straight line objects.

Does it work and if so , can anybody give a good example?.


Here is a link to a posting a few years ago by Henry, the lofted version will work, but I am buggered if I can get the extruded methods to work.

Link to A previous post on the forum http://forums.turbocad.com/index.php/topic,1638.0.html


Any help appreciated!.

Logged


* October 01, 2013, 06:38:21 AM
#1
It's impossible to hob or cut the tooth form of a true bevel gear, helical or straight, in real-world engineering, so the ability to do it in CAD  is kind of arbitrary.  I've got access to another program that twists solids after their production, so I coned a gear profile in TC, exchanged the .sat solid and twisted i

Logged


* October 01, 2013, 10:28:52 AM
#2
I agree about it being unlikely to get a truly accurate representation Murray, I am just puzzled how draft angle works. I have been using Henry's loft with two profiles method and then twisting one of the profiles with minimize twist unchecked. I did manage to do it in Moi3d which was with the same twist methodology that you have used. I wish we had Michael Gibson(Moi3d) on board at IMSI, with his ingenious routines for some of the more exotic forms.


Logged


* October 01, 2013, 01:50:30 PM
#3
Looking at Murray's picture made me think that the Twisted Extrude tool might work. Tweak the various options; such as Twist Continuity. Unfortunately, the Draft Angle only affects the teeth, from what I see. I don't see any way to make one end smaller.

I can twist a Cone after it's made. Select it, then use "Select Profile" and rotate. It doesn't look as nice as Murray's though.
« Last Edit: October 01, 2013, 01:53:16 PM by John R »

Logged
John R.

V17—V21, 2015—2019
Designer, Deluxe, (Professional, Expert, Basic), Platinum
RedSDK enabled
Windows 10 Pro (1903), 64-bit


* October 01, 2013, 04:30:15 PM
#4
Geez you're knowledgeable, John.  I thought it might be possible in TC, because the other program that I used has very similar spec to TC, ACIS, lightworks, Dcubed constraints, but you know TC inside and out.
Michael, draft is conventionally the minimum angle from normal to the parting line of a die or mold, and it's usually dealt with on an individual basis for each face (had to resist "face-by-face basis" there.  But I had to share.  So conflicted.).  I don't know how successful expecting it to deal with multiple, opposed, compound-curved faces is likely to be.  Perhaps it could be used to do what you want on one tooth, radial copies for the remainder.
I'm not convinced that programs like MoI are compatible with programs like TC.  It's got neat routines like flow and twist, but they're individual routines that don't sit comfortably with feature-based, or history-based CAD.  Because it's a surfacing program, it can generate self-intersecting objects without blinking, while that's a prime no-no for kernels that're predicated on manufacturability.  After-the-fact surface replacement, which is the arrangement that MoI-with-Alibre worked along, isn't like trying to build it into the CAD app.
« Last Edit: October 01, 2013, 04:43:38 PM by murray dickinson »

Logged


* October 01, 2013, 07:16:00 PM
#5
"I can twist a Cone after it's made. Select it, then use "Select Profile" and rotate. It doesn't look as nice as Murray's though."

Excellent.

Henry H

Logged


* October 01, 2013, 08:35:29 PM
#6
It occurs to me that it should be possible to set up the twisted section as a face-to-face loft between two coned gears.  It'd probably work better using fewer teeth than a full gear - I'm thinking more twist range - and radial copies added.
« Last Edit: October 01, 2013, 08:38:33 PM by murray dickinson »

Logged


* October 02, 2013, 10:03:34 AM
#7
Michael, draft is conventionally the minimum angle from normal to the parting line of a die or mold, and it's usually dealt with on an individual basis for each face (had to resist "face-by-face basis" there.  But I had to share.  So conflicted.).

Had an idea it might not mean what I originally thought, but don't do any casting work, so never really came up(makes sense now).

The single tooth idea is a good idea, if it produces a good profile, I remember using it on a segment of worm gear for a sextant a couple of years ago, also saves a lot of time with the tooth as a block and radial copied.

What is more important for the drawing is a good acceptable visual representation, rather than a dimensionally perfect profile, but both would be nice!.

I might try the ordinary bevel gear rotated in steps, and take sections at various elevations on the vertical axis, then try a loft with guidelines, although not sure if the guidelines method works that well.

Many thanks for the feedback everyone!.

Logged


* October 02, 2013, 05:30:52 PM
#8
I might try the ordinary bevel gear rotated in steps, and take sections at various elevations on the vertical axis, then try a loft with guidelines, although not sure if the guidelines method works that well.

Try John's method: Insert a "Gear Contour"; explode it to a Polyline; use it as the 2D Base for a Cone (with "Cut Cone as Lofting" enabled). After the cone has been created, select it and click "Select Profile" on the IBar. Rotate the profile to create the desired twist. Works great.

Henry H

Logged


* October 02, 2013, 10:44:52 PM
#9
It does work great, but the result's a skew bevel rather than helical. 

Logged


October 03, 2013, 12:21:12 AM
#10
It does work great, but the result's a skew bevel rather than helical. 

IMHO that's sort of half the story. It's a skew bevel gear, the pinion in a skew bevel gear set is a plain ordinary bevel wheel offset from the gear centreline, (hence the need to skew the gears teeth). Their advantage is they can be cut on a milling machine. Older car differentials had this type of gearing. You need at least one intermediate rotated and scaled profile to achieve a spiral.

A spiral bevel gear set has the pinion on the gear centreline. If you offset a spiral bevel gear pinion you end up with a hypoid gear set.

Hope that helps

Best Regards
Bob
« Last Edit: October 03, 2013, 12:23:41 AM by Bobich »

Logged
TC 14 Deluxe, TC 19 Pro
Windows 7, 64bit 8GB Ram


* October 03, 2013, 08:40:25 AM
#11
It does work great, but the result's a skew bevel rather than helical.

Mike wrote a couple of days ago, "What is more important for the drawing is a good acceptable visual representation, rather than a dimensionally perfect profile, but both would be nice!".

Henry H

Logged


* October 03, 2013, 09:31:22 AM
#12
I can twist a Cone after it's made. Select it, then use "Select Profile" and rotate. It doesn't look as nice as Murray's though.


I missed this part John although Henry picked it up, thought it was a cone with two lofts, did not know you could edit the profile of a cone after creation.


Logged


* October 03, 2013, 03:48:34 PM
#13
As long as it stays in its "primitive" state or as part of the Part Tree history, the profiles can be edited. Once it becomes an ACIS Solid, there's no way to twist it. Facet Edit won't do it.

Logged
John R.

V17—V21, 2015—2019
Designer, Deluxe, (Professional, Expert, Basic), Platinum
RedSDK enabled
Windows 10 Pro (1903), 64-bit


* October 03, 2013, 05:39:13 PM
#14
It does work great, but the result's a skew bevel rather than helical.

Mike wrote a couple of days ago, "What is more important for the drawing is a good acceptable visual representation, rather than a dimensionally perfect profile, but both would be nice!".

Henry H

I submit that "important", "good", "acceptable" and "nice" are underconstrained parameters that can only be reconciled to an individual degree of pedantry, Yer Honour.

Logged


* October 04, 2013, 01:53:59 PM
#15
After saying what I said Murray, I have just created an even more complex version that takes two profiles from a spur gear, and loft them perpendicular to the side of the cone creating my bevel gear, which looks far better when meshing for visual effect, not a helical version though.

When gears start to be manufactured right out of the drawing to cnc, it will become necessary to create dimensionally correct gears, but until then, the visually more appealing and best realistic timeframes for creating will take precedence.

For me especially as I create a lot of my models for my artwork, artistic license can make the image more appealing.

Does anyone have any good info on "Octoid" profile development, which is supposed to be the closest to the true tooth form.

Logged


* October 04, 2013, 06:16:13 PM
#16
The contact points of cylindrical gear teeth with involute profile don't slide against each other at all, the line of action is straight, tangential to both gears' pitch circle.  Bevel gears would be the same if it were possible to create perfect conical involutes.  But involutes are generated by helical cutters, hobs, which must be cylindrical.   So, a bevel gear is generated with a straight-flanked cutter that moves as it cuts each tooth in such a way that the line of action traverses a figure-eight path.  There's a nice graphical explanation here, on page 1: http://pdf.directindustry.com/pdf/quality-transmission-components/bevel-gearing/11708-82731.html    It approximates an involute except that, instead of there being a theoretical contact line across the whole of the teeth of cylindrical gears, there's a point of contact only, that moves in and out across the teeth, and that's the path that wears first and worst.  Bevels are normally cut and lapped as pairs, they're always going to be noisier than spur gears, wear faster, and have to be replaced together if the set's performance is to be maintained. 

Logged