Straight Hob Tooth Face location

[Paul FallertParticipant @paulfallert28101]

This question is about making a straight hob. The hob will be used to cut a 14 tooth pinion which has a 30 deg pressure angle and 44DP. The diameter of the pinion is .363" and it will be .860" in length to fit a friend's broken tool. I provide this detail in the event that it provides some relevance to the question.

Having researched this topic in WPS17-Ivan Law's book on Gears and Gear Cutting, MEW #131, Loop's Straight Hobbing article in StrictlyIC-Aug,1994, JohnEdgar on GearHobbing, p6, several websites and YouTube videos, I have one remaining question about making a straight (as opposed to spiral) hob. The straight hob method appears to be the easiest process to get to an accurate working tool, however it may be the slowest in actually making the pinion. Given the unusual PA and DP, purchase of a gear cutter under the circumstances makes little economic sense.

Should the face of the teeth be on the center line of the hob or should they be offset from the center line to cut a steel pinion? And if offset, by how much?

Paul

—BTW, Harold Hall's latest MEW article on pinions just happened to appear on schedule and in it, he has again provided a timely solution, in this case how to go about holding the end of a pinion blank. Thank's Harold.

[Paul FallertParticipant @paulfallert28101]

On center or right of center line?

[Anonymous]

On the hob I am making to cut a worm gear in cast iron the front face of the teeth will be on a radial, same as for standard involute cutters.

Regards,

Andrew

[Sub MandrelParticipant @submandrel]

Using Ivan Law's method errors are introduced by both the angle of the buttons used to form the cutters (reduces their effective diameter) and that the cutter edges are not truly radial (increases it). Neither of these errors is discussed, but after worrying about it, I decided they probably cancel out within the bounds of what matters as the final result works.

Neil

[Paul FallertParticipant @paulfallert28101]

Neil:

You are making me more convinced that I should avoid the button method. The use of buttons and related precise measurements to prepare the form cutter has a lot (to me anyway) of pitfalls, which is why I decided to instead use the Straight Hobbing Method.

Preparing the Straight Hob is easy, other than where to locate the face of the teeth. In contrast to a Spiral Hob, you don't need change gears or even a screwcutting lathe. For a pinion of 30 deg Pressure Angle, a regular 60 deg threading tool can be used. The involute shape of the tooth is formed on the pinion (or gear) by the rolling action [my term.] of the hob's lands as they progressively form the teeth so long as you make relatively shallow cuts with the hob (thus, lengthening the time expended on the making of the pinion).

Thanks for responding to my request.

Paul

[Paul FallertParticipant @paulfallert28101]

Andrew:

Restatement of the question:

Looking at the end of the hob, if I were to draw a "+" sign on the face of the hob with the center of the "+" sign crossing through the center of the hob and the horizontal bar of the "+" sign on the horizon, should the face of the first tooth formed in the hob be on this vertical line? or set back from the line by .010" to .062" as I have seen in some published articles?

Or, are you saying to place the face of the tooth on a radial line would be the same as placing it on this vertical line?

Paul

[Anonymous]

I'd put the face of the first tooth on the vertical line of the cross. If the centre of the cross is on the centre of the hob then by definition the arms of the cross are radials, so we mean the same thing. I assume that the setting back of the face of the tooth is an attempt to get some rake? In my experience involute gear cutters work fine with zero rake, even in EN8.

I'm a bit confused by the terminology of spiral and straight hobs? To my mind a hob always has the teeth arranged on a helix. That way, as the hob rotates, and the gear blank rotates in sympathy, the straight sided teeth on the hob automatically generate the involute form. The gashes in the hob to create separate teeth can be straight or helical, but that's different. If by straight hob you mean that the teeth are of the correct form, but are arranged in a series of parallel circles I don't see how that is going to generate the teeth on the pinion, unless you manually index the gear blank and move the hob sideways at each pass.

Regards,

Andrew

[Paul FallertParticipant @paulfallert28101]

Andrew:

Thank you for your complete and most helpful reply.

Straight Hobbing is new to me, too. The hob is prepared without a helix ! Each land is perpendicular to the axis of the hob.

The term "straight hob", unfortunately. does not search well, but some books on gear hobbing make reference to the process, and Ivan Law makes a curt reference to the process in WPS#17.

You should have at least 5 lands for the method to work (even 3 will work on a small diameter gear) to create an involute tooth form. You cannot, for example have but one land, for it would not generate an involute tooth form. On a large gear, you need 7 lands.

As you surmised, you must manually index the gear blank for each tooth and you must make shallow passes of the hob cutter into the gear blank. The concept is for the lands above and below the central land to gradually remove a bit of metal with each index pass. The central land is the middle land (#3 in a 5 land hob) and it must be cutting on the centerline of the gear blank.

Straight hobs are EASY to make. The lathe tools are easy to grind accurately and quickly, unlike buttons. The process automatically corrects for all but the most egregious errors. It is best to file a thin sheet metal template for guidance in grinding the angle of the lathe tool to the pressure angle (PA), as in my case a 30 deg PA requires a lathe tool to cut a 60 deg groove between each land. Land spacing equals the Circular Pitch (CP), but to cut the grooves forming the hob lands since they are parallel, you only need to use your lead screw micrometer dial or lacking a lead screw micrometer or even lacking a lead screw, a dial indicator poised against the side of the saddle will accurately measure the distance moved between the lands. A screw cutting lathe is net required to make a Straight Hob, even a wood lathe could do it. The diameter of the material used to make the hob is not important, but of course you must take the diameter into consideration when deciding where to form the teeth. The number of teeth is not important. Each tooth must be relieved or it will not cut. I am thinking that rake might be helpful if the mill used to hob the gear blank/pinion is not substantial. The entire process could be done on a lathe with a milling attachment.

I suspect this is very old technology and may pre-date the screw cutting lathe, maybe going back to early clock and watch makers who needed a way to make gears and pinions more quickly than hand filing.

Paul

[Anonymous]

Of course, you don't need to move the hob, just index the gear blank. So one ends up with the 'involute' tooth form actually consisting of a series of straight lines, the number of lines being dependent upon the number of teeth in the gear. I'm not quite sure why one has to take shallow cuts with the hob? Wouldn't full depth in one go, and indexing as described suffice?

It's a moot point as to which of a straight hob or a single point tool based on 'buttons' would give the better approximation to the correct involute curve.

Regards,

Andrew

[BazyleParticipant @bazyle]

I think you move the hob sideways as you rotate the gear – in sync – as the hob is a rack equivalent.

The purpose of the question presumably is whether having the face off-radial improves the cutting ability by providing sort of top rake. I think it might but is probably not essential. The cutting action is a bit like plunge cutting a screw thread with the tool moving and the work fixed. Do you make your screwcutting tool without top rake? Ok for brass, better to have some for steel but still works.

[Sub MandrelParticipant @submandrel]

As the number of 'lands'/'cuts' increases the straight hob exponentially approaches a true involute. I have read that the 5/7 cuts Paul mentions will [produce a gear within tolerance. My experience is that the button method works, I don't know if the increased cutting speed of a hob over a single point cutter makes a big difference once the fiddling to get the multiple passes right is taken into account.

I find the slow process of cutting a gear in steel with a single point cutter very zen, anyway.

Neil

[John HinkleyParticipant @johnhinkley26699]

First of all, let me say that I know less than nothing about gear cutting – so what I'm about to write could well be way off the mark, but here goes anyway!

I have read and re-read this thread until my brain hurts. Still don't get it! Sorry to keep plugging these guys, but have you tried CTC Tools in Hong Kong? They advertise involute gear cutters for very little money, all sizes from what I can see. Would one of these not fulfil your requirement, Paul?

That's my 2p. Back to swat up on gear cutting.

John

Edited By John Hinkley on 02/10/2013 16:34:06

[Keith LongParticipant @keithlong89920]

You don't have to move the hob but you do have to move the gear blank sideways in relation to the hob as well as indexing it rotationally. Which you move is up to you but if you're doing this on most normal machines I would guess that moving the gear blank would be easier. In effect the gear being cut is meshed with a rack – the hob – that just happens to be cutting (generating) the teeth on the gear, so if you rotate the gear to take the next cut then the gear would roll along the rack getting a linear displacement. If you don't move the gear along in the end you'll just wind up machining all the teeth off.

Keith

[Anonymous]

Posted by Stub Mandrel on 02/10/2013 15:48:06:

As the number of 'lands'/'cuts' increases the straight hob exponentially approaches a true involute.

That can't possibly be true? Let's take the example of, say, a 14 tooth pinion as described by the OP. If we have one land on the straight hob it will cut a straight sided gap, not a good approximation. If we add two more lands, ie, 3 in total, then the outer two will start cutting gaps before the central land sees the gap. This is a better approximation. If we add another two lands the approximation gets better, and so on. However, the gear blank also falls away from the hob, so at some point adding two more lands will have no effect, as they will never contact the blank. So adding more and more lands does not increase the accuracy, exponential or otherwise, QED.

However, if we increase the number of teeth on the gear to be cut, then the blank falls away more slowly, so that more lands on the hob are in use, and the approximation gets better. So the accuracy actually increases with the number of teeth on the gear, not directly with the number of lands on the hob. But, there's always a but (!), that isn't strictly true either. That 'rule' implies that to cut a rack, a gear of infinite radius and thus an infinite number of teeth, we'd need a hob with an infinite number of lands. But in fact to cut a perfect rack we only need one land.

As in the best text books, it is left as an exercise for the reader to determine what the optimum number of lands is for a given number of teeth.

Regards,

Andrew

[Anonymous]

Posted by Keith Long on 02/10/2013 17:12:25:

You don't have to move the hob but you do have to move the gear blank sideways in relation to the hob as well as indexing it rotationally. Which you move is up to you but if you're doing this on most normal machines I would guess that moving the gear blank would be easier. In effect the gear being cut is meshed with a rack – the hob – that just happens to be cutting (generating) the teeth on the gear, so if you rotate the gear to take the next cut then the gear would roll along the rack getting a linear displacement. If you don't move the gear along in the end you'll just wind up machining all the teeth off.

Keith

There is no need to move the hob if you index the gear by one exactly tooth space each time. If one indexes the gear by an arbitrary amount and also move the hob to compensate, it'd take for nearly ever to cut the gear. Since the hob is not infinitely long you'd have to reset it's position every so often as well. It shouldn't be that difficult to calculate the rotational and linear moves, but the chances of getting every one right is rather small, at least for me, as I'd lose interest, and concentration, pretty quickly.

Regards,

Andrew

[Paul FallertParticipant @paulfallert28101]

After the hob has been finished, the part of the process that I have not elaborated on is as follows [using vertical milling machine nomenclature for simplicity]

Position the gear blank on the table in an indexing fixture with it's axis aligned parallel to the length of the mill table. Lock the indexer spindle from rotation.

Place the hob in the vertical spindle. Align the Straight Hob's central land to the centerline of the gear blank.

Start the hob rotating and move the table's Y axis (the narrow table axis) so that rotating hob moves closer to the gear blank. Lock the mill table's Y axis. Now the gear blank is locked from rotation and the Y axis is locked, fixing the depth of cut.

Cause the table to move past the hob. The table would move on its X (long) axis, normally that would be a table moving from the Left to Right, assuming the gear blank had been positioned to the left of the spindle.

Return the table to its starting position. Do not climb mill with the hob.

Index the gear blank by one tooth. Unlock the Y axis and bring the hob closer to the gear blank so as to be able to make a deeper cut. Tighten the Y axis table lock. Make the next cut, again moving the table L-R. Repeat until all of the spaces between the gear blank's teeth have been cut.

Note: It should now be clear that you do NOT simultaneously index the gear blank while the hob is cutting, as that is what happens with a spiral hob, not applicable to a Straight Hob.

Paul

[Sub MandrelParticipant @submandrel]

HI Andrew,

Think instead of rotating the gear through exactly half a pitch along a rack. The point of contact will, in theory, move from the root to the tip of the gear. This is what would happen with an infinite number of passes of the hob.

Intuitively (and I'm sure it is correct) two passes will be much better than 1, 3 a bit better than 2, but by the time you get to 8 the improvements will be tiny.

Neil

[Paul FallertParticipant @paulfallert28101]

For those of you still having trouble with this concept, MEW #131, page 26 does have pictures of how the progressive cutting of the teeth do form an involute gear tooth.

If you follow that article, you will find that you can "easily" produce workable gears.

To answer the suggestion about purchasing a spiral hob from say CDC in Hong Kong, the choice of available spiral hobs is limited. In my case, I need a 30 deg Pressure Angle with a 44DP for a repair part for a friend and not wishing to buy an odd tool that will only ever be used once.

Once you understand the process, with the Straight Hobbing technique, you can produce gears with unusual diametral ptiches and pressure angles at virtually little or no cost. You just need a piece of drill rod and the drill rod can be as small as 12mm. And one individual used a piece of tool steel from an old power chisel bit, first softening the steel, making the hob and then re-hardening the steel.

Maybe this thread would make a subject for an illustrated series in a future magazine article. (just kidding).

Paul

[BazyleParticipant @bazyle]

Isn't this all the same as shaper cutting gears? A topic I will leave you to search as it was covered on another forum a couple of months ago.

[Anonymous]

Posted by Bazyle on 02/10/2013 20:46:55:

Isn't this all the same as shaper cutting gears?

Precisely; it's exactly how a Sunderland gear planer works. A short length of rack is moved back and forth, and between each stroke the gear blank is rotated slightly, and the rack moved slightly to keep in step. Every few strokes the rack needs to be reset as it is of finite length.

The issue is how much the gear blank is indexed between cuts. If I understand the OP correctly he is indexing the gear blank exactly one tooth position between cuts, and therefore the hob does not need to move. In this case my analysis holds. If the gear blank is rotated by less than a tooth pitch then the hob also needs to be moved an appropriate amount, in the same way that a rack engaged with the gear would move. In this case the accuracy of the generated involute depends upon the size of the step. It does not depend upon the number of lands on the hob; that just determines how often the hob needs to be reset.

Regards,

Andrew

[Sub MandrelParticipant @submandrel]

I read it as the 'lands' being the number of flats that make up any gear profile. There have been some cross-purposes here!

On re-reading, it is the number of 'teeth' on the hob that have some engagement with the gear, and yes, the gear and hob are not moved sideways with this method.

Neil

[Anonymous]

Posted by Stub Mandrel on 03/10/2013 18:29:46:

I read it as the 'lands' being the number of flats that make up any gear profile. There have been some cross-purposes here!

Indeed, but at least it only involves lands, better than being all at sea – Andrew

Addendum: It's been an interesting discussion, but if I were doing the job I'd use a home made cutter similar to that made by the 'button' method. However, I'd make the cutter on a four axis CNC mill, so the profile will be 'exact' within the limits of the mill, and could be multi tooth.

[Paul FallertParticipant @paulfallert28101]

Finally found a short treatise on the Straight Hobbing Method.

The following URL explains the "Straight Hobbing" method. Be sure to look to the left side of the referenced web page to get access to the discrete sections, e.g. See pages titled "Involute Gears", "Method", "Cutter", "Cutting".

http://homepage.ntlworld.com/peter_harrison/workshop/gearcutting/index.htm

That site provides step-by-step method and pictures. Note: Neither "Straight Hobbing" nor "lands" are used in that article.

Thanks to those who provided all of the useful information.

Paul

 

 

Edited By Paul Fallert on 04/10/2013 21:46:25

[Paul FallertParticipant @paulfallert28101]

The URL link should have been encoded as a "Link".

**LINK**

Paul

Edited By Paul Fallert on 04/10/2013 21:48:22

[John Stevenson 1Participant @johnstevenson1]

Tony Jeffree also did an article recently in MEW on the same method.

Not got my index handy but some kind soul will chip in with what number it was.

[Author]

Posts