PDA

View Full Version : Involute shaft



John Stevenson
05-22-2010, 06:00 PM
Bear with me this could take a week to do from start to finish as will be explained in the post but I hope to show it with pictures and text.

Got a job in last week to do some short shafts with an involute spline on the end, even got a drawing that told me the OD, how many splines and how long- wow - but nothing else.
0.650" OD, 0.875" long and 12 splines.

Involute splines are just gears but only half the depth and usually have a 30 degree pressure angle and occasionally 25 degrees.
Because of this gear calculations work the same - to an extent.
splines are usually denoted by both sizes of the make up, like 16/32 DP.
This means the teeth are 16DP but the depth is equal to 32 DP

To calculate the DP you take the number of teeth, add one and divide it by the outer diameter, for a normal gear it's number of teeth plus 2 and divide. splines are plus 1 because the depth is half of a normal gear.
So in out case 12 + 1 = 13 / 0.650 =20 DP so this is a 20/40 DP spline.

Run these figures thru the gear generator and we get this.

http://www.stevenson-engineers.co.uk/files/12t20_40dp.jpg

Note 12 teeth, 20DP, and the important bit the depth factor has been entered as .5 to allow for the depth of a 40DP spline / gear.
This then gives us a true involute of the complete spline which can be saved as a DXF file.

Once this is brought into CAD and zoomed in you can draw two straight lines, one from the root of the tooth to halfway up the involute and another from there to the outer tip.

Now if two new lines are drawn midway on each line and at 90 degrees to the existing lines they will converge on the centre of a circle that passes thru all three points, standard geometry 101.

Draw a circle to check and repeat on the other side.

http://www.stevenson-engineers.co.uk/files/12teeth20_40dp.jpg

This gives us four important pieces of information to make a cutter, the size of the circles, 0.277" and the distance apart 0.300", the width of the gear cutter blank at 0.119" and the depth the buttons need to be infed, 0.051"

Next job is to make what is usually termed a button cutter.

http://www.stevenson-engineers.co.uk/files/button%20tool.jpg

Just a piece of flat bar with two holes drilled in at 0.300" centres and tilted at 5 to 7 degrees to accept two top hat buttons made of Silver steel or drill rod , hardened and tempered to light straw for cutting steel.
These are then surface ground on the top face to sharpen them and give a cutting angle.

This button tool is used to prepare the cutter blank, more of which, with pics in a later post.

.

millwrong
05-22-2010, 09:53 PM
I don't think that I've ever seen a button cutter-in 3D or in action! I for one am eagerly waiting for more spline teasing!

MotorradMike
05-22-2010, 10:10 PM
Thanks for posting this.

I'm not likely to ever need to do anything like it it myself but it's very interesting all the same and I really appreciate that you have the patience to share your knowledge.

doctor demo
05-22-2010, 11:07 PM
you can draw two straight lines, one from the root of the tooth to halfway up the involute and another from there to the outer tip.

Now if two new lines are drawn midway on each line and at 90 degrees to the existing lines they will converge on the centre of a circle that passes thru all three points, standard geometry 101.
.
John, You lost Me . I only made it to geometry 100.5 not 101:)

Steve

PS and please don't call Me a clumsy bass turd.

rollin45
05-22-2010, 11:19 PM
Great Stuff Sir John!!

I'll most likely never use it, but it sure is interesting, I also want to say thanks for your time and efforts in posting this.

best

rollin'

Jpfalt
05-22-2010, 11:26 PM
Where i used to work we had JIS splines that were the half depth you described, but the PD on the spline was actually at the OD per JIS standard as well as using a 20 dsegree pressure angle.

Is that possibly the case with your splines?

If so, it makes the spline thicker at the top of the spline and the curvature less on the spline faces.

TGTool
05-22-2010, 11:35 PM
Where i used to work we had JIS splines that were the half depth you described, but the PD on the spline was actually at the OD per JIS standard as well as using a 20 dsegree pressure angle.

Is that possibly the case with your splines?

If so, it makes the spline thicker at the top of the spline and the curvature less on the spline faces.

I think when John makes the shafts it will be up to the customer to adapt their parts. :D

Black_Moons
05-23-2010, 12:31 AM
Im intrested in this 'button' method.. However a nagging feeling has struck me...

How did you align circles to involute curves? Should'nt they be.. involute not circular?
http://upload.wikimedia.org/wikipedia/commons/thumb/8/88/Animated_involute_of_circle.gif/200px-Animated_involute_of_circle.gif

boslab
05-23-2010, 12:42 AM
I dont think i get the spline generation but i like the gif, cool animation
in time i may yet understand
mark

dp
05-23-2010, 12:45 AM
Im intrested in this 'button' method.. However a nagging feeling has struck me...


The button method is well described in Ivan Law's little red book of gears - I consider it essential reading.

http://www.amazon.com/Gears-Gear-Cutting-Ivan-Law/dp/0852429118

Here's Johns' own version: http://www.metalwebnews.com/howto/gear/gear1.html

And once you have the gear cutter you need a reliever: http://www.youtube.com/watch?v=kJ8kyC_bpHs

NzOldun
05-23-2010, 01:07 AM
Bear with me this could take a week to do from start to finish as will be explained in the post but I hope to show it with pictures and text.

Got a job in last week to do some short shafts with an involute spline on the end, even got a drawing that told me the OD, how many splines and how long- wow - but nothing else.
0.650" OD, 0.875" long and 12 splines.

Involute splines are just gears but only half the depth and usually have a 30 degree pressure angle and occasionally 25 degrees.
Because of this gear calculations work the same - to an extent.
splines are usually denoted by both sizes of the make up, like 16/32 DP.
This means the teeth are 16DP but the depth is equal to 32 DP

To calculate the DP you take the number of teeth, add one and divide it by the outer diameter, for a normal gear it's number of teeth plus 2 and divide. splines are plus 1 because the depth is half of a normal gear.
So in out case 12 + 1 = 13 / 0.650 =20 DP so this is a 20/40 DP spline.

Run these figures thru the gear generator and we get this.

http://www.stevenson-engineers.co.uk/files/12t20_40dp.jpg

Note 12 teeth, 20DP, and the important bit the depth factor has been entered as .5 to allow for the depth of a 40DP spline / gear.
This then gives us a true involute of the complete spline which can be saved as a DXF file.

Once this is brought into CAD and zoomed in you can draw two straight lines, one from the root of the tooth to halfway up the involute and another from there to the outer tip.

Now if two new lines are drawn midway on each line and at 90 degrees to the existing lines they will converge on the centre of a circle that passes thru all three points, standard geometry 101.

Draw a circle to check and repeat on the other side.

http://www.stevenson-engineers.co.uk/files/12teeth20_40dp.jpg

This gives us four important pieces of information to make a cutter, the size of the circles, 0.277" and the distance apart 0.300", the width of the gear cutter blank at 0.119" and the depth the buttons need to be infed, 0.051"

Next job is to make what is usually termed a button cutter.

http://www.stevenson-engineers.co.uk/files/button%20tool.jpg

Just a piece of flat bar with two holes drilled in at 0.300" centres and tilted at 5 to 7 degrees to accept two top hat buttons made of Silver steel or drill rod , hardened and tempered to light straw for cutting steel.
These are then surface ground on the top face to sharpen them and give a cutting angle.

This button tool is used to prepare the cutter blank, more of which, with pics in a later post.

.
John,

I have seen your video on You tube, cutting a spiral bevel gear on your X3 under CNC.

Just a thought. Using your CNC, it would appear possibe to make a rack shaped cutter and generate the spline teeth by using the the rotary table in the horizontal position, and running the main table back and forth in the manner of a Sunderland gear planer, but moving the workpiece rather tah the cutter? Thinking about it, it would be vital to lock the main spindle, to prevent rotation, if this is where the cutter was mounted!

Just a thought

Cheers
NzOldun

Paul Alciatore
05-23-2010, 02:11 AM
The button method is well described in Ivan Law's little red book of gears - I consider it essential reading.

http://www.amazon.com/Gears-Gear-Cutting-Ivan-Law/dp/0852429118

Here's Johns' own version: http://www.metalwebnews.com/howto/gear/gear1.html

And once you have the gear cutter you need a reliever: http://www.youtube.com/watch?v=kJ8kyC_bpHs

The involute is a curve with a constantly changing curvature (radius). A circular button will have a single radius.

So, the button method produces a cutter that is an approximation of the involute form, not exact. But it can be quite close, usually within a thousanth or less.

dp
05-23-2010, 02:27 AM
The involute is a curve with a constantly changing curvature (radius). A circular button will have a single radius.

So, the button method produces a cutter that is an approximation of the involute form, not exact. But it can be quite close, usually within a thousanth or less.

That is correct. In fact though, because the buttons are tipped at 5 or so, they produce elliptical shapes. Some years back John produced CAD drawings of the button method profile overlaid on a generated involute for a variety of pressure angles/tooth counts and the differences were withing the error produced by using commercial cutters over the entire tooth count range.

Black_Moons
05-23-2010, 05:26 AM
ahh tilting the button, that would produce a very close approximation, cool.

John Stevenson
05-23-2010, 06:37 AM
Might be a couple of days before the next instalment, the tool and cutter blank but some answers first in no order.

The circle IS only an approximation as Black Moons has said but the arc length of this spline is only 0.066" long out of a circle who's circumference is 0.870" so we are talking about a 13th part of a circle.
To be honest I don't have anything that could measure this error.

Finding the centre of an arc or circle for people who got distracted doing Geometry 100.5 :rolleyes:

http://www.stevenson-engineers.co.uk/files/CIRCLECENTRE.jpg

Draw two lines that touch on the circle at each end, then draw another line at 90 degrees to this from the centre.
where these to lines meet is the arc or circle centre.
Because the spline is so small it's hard to see the original work without zooming in.

NzOldun,
That is doable and is worth giving some thought, however now I have started this I will have to continue and as I reckon the job will be a repeat it's worth it to have the cutter.

I'll give it some thought and if I do I'll do another unrelated thread on it.

JPFalt,
I did input various pressure angles but as soon as I got down to 25 degrees the figures were way off. In all fairness I have never seen JIS splines and I'm under the impression they are for newer automotive designs which I won't see for 40 years [ as if ] ;)

lazlo
05-23-2010, 10:29 AM
i like the gif, cool animation

Wikipedia knows all :p


Some years back John produced CAD drawings of the button method profile overlaid on a generated involute for a variety of pressure angles/tooth counts and the differences were withing the error produced by using commercial cutters over the entire tooth count range.

John's CAD drawing doesn't show that. It shows that the difference between the button method and a true involute is very small, for that particular cutter.

Involute cutters are only accurate for a specific number of teeth. On a commercial involute cutter, they're only accurate for the lowest number of teeth in the range. They're intended for job shops, not for production use.

When you get into commercial gears, they have AGMA ratings, which are like ABEC ratings for bearings. Commercial gears are usually hobbed, which produces a perfect involute form for any number of teeth. Higher quality gears are shaved after hobbing for even more accuracy and surface finish:

http://www.gears-manufacturers.com/agma-standards.html

http://i164.photobucket.com/albums/u15/rtgeorge_album/AGMAratings.png

John Stevenson
05-23-2010, 10:49 AM
When you get into commercial gears, they have AGMA ratings, which are like ABEC ratings for bearings.

Yabut that only applies to American Gear Manufacturers Association, AGMA, and us Brits are exempt. :rolleyes: :D ;) :p

lazlo
05-23-2010, 10:59 AM
Yabut that only applies to American Gear Manufacturers Association, AGMA, and us Brits are exempt. :rolleyes: :D :p

Nah, like ABEC (American Bearing Engineers Committee) -- you just have to translate into those funny units. Oh wait, you're British -- you still use Imperial :p

By the way John -- nice job as usual! Looking forward to the pictures of the spline shaft!

dp
05-23-2010, 11:34 AM
John's CAD drawing doesn't show that. It shows that the difference between the button method and a true involute is very small, for that particular cutter.

You repeated, with graphs and tables, what I said.

lazlo
05-23-2010, 11:39 AM
Re-read what I wrote.

John's CAD drawing (which he still has on his web page), shows the difference between the button arc and a perfect involute. He does not plot a commercial involute cutter, measure the error associated with a commercial involute cutter, or show that the button method is within the error of a commercial cutter.

Tony Ennis
05-23-2010, 11:56 AM
JS's stuff tends to be 'best by test.' If he's good with a certain amount of error between theoretical involutes and the profiles cut by his canted circular cutters then I'm good with it too.

I was under the (mistaken?) assumption that the generally approximate involutes provided by standard cutters were good enough to use without modification. So the first education here is that some people clean them up a little bit. The second is that it doesn't look too hard, though the reality is it would probably take me days to make the tool he describes.

dp
05-23-2010, 12:08 PM
Re-read what I wrote.

John's CAD drawing (which he still has on his web page), shows the difference between the button arc and a perfect involute. He does not plot a commercial involute cutter, measure the error associated with a commercial involute cutter, or show that the button method is within the error of a commercial cutter.

I was thinking in the thread this comes from he'd done some drawings as I described: http://bbs.homeshopmachinist.net/showpost.php?p=397212&postcount=21

It may have been someone else's drawings, but the message remains - the button method error is there but not significant.

The images I found have no context but are still interesting:

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/12T_16DP.BMP
http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/26T%20true.BMP
http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/30T_14.5.BMP
http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/30_32_35.BMP
http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/35T_16DP.BMP

mochinist
05-23-2010, 12:23 PM
It would be cool if there was a thread option for project or how to threads that could only be posted in by the thread author until the project/how to was considered done. Then everyone could ask questions and all the know it alls could have a blast arguing about what they actually said and post graphs and gifs.

dp
05-23-2010, 12:30 PM
It would be cool if there was a thread option for project or how to threads that could only be posted in by the thread author until the project/how to was considered done. Then everyone could ask questions and all the know it alls could have a blast arguing about what they actually said and post graphs and gifs.

That's called a blog - they have them at PM.

mochinist
05-23-2010, 12:36 PM
That's called a blog - they have them at PM.yeah a blog is a little different than what I am thinking, I like the thread format but it would be nice to see all of the authors post in order and then when they are done, questions, praise and arguments could commence. Anyways now I'm doing to this thread what I dont like to see:p

Jpfalt
05-23-2010, 01:12 PM
I haven't noticed anyone mentioning using a rack gear form to generate the involute. It involves some mechanical setup, but always generates a true involute. All the methods described so far approximate an involute that is close enough for any practical standard.

The method I mean by rack gear is to set up a cutter or broach in the form of a gear rack with depth and pressure angle as needed and pitch line known. The shape is flat topped, straight tooth sides and flat bottom or whatever relief you want built in. The blank is advanced down into the rack and the rack reciprocated along the axis of the gear until the desired pitch diameter of the gear reaches the pitch line of the rack profile cutter.

The rack cutter is then moved sideways while reciprocating while the gear is rotated so that the pitch diameter of the gear and the pitch line of the rack cutter move the same distance. On shapers this was done by using a spring steel strip on a cylindrical drum that was part of the gear blank holder. as the blank is moved laterally, the steel strap rotates the arbor.

The end result is an involute as accurate as the cutting tool and fixture were originally made. I've also seen involute teeth generated on a shaper using a single point tool as a single tooth rack gear and with the spring and drum to generate one tooth gullet at s time. The nice thing is that the generating tool is so easy to grind to profile.

I've been thinking about getting some rack gear sections diamond plated by Amples to use as a truing file for gear teeth.

Cad wise, it's pretty easy to generate an involute in sketch mode in AutoCad or Solidworks by doing the same process electronically.

Carld
05-23-2010, 01:28 PM
What I am interested in seeing is how this "button cutter" is going to cut the spline teeth. I don't see how it will work and am looking forward to seeing it in action.

dp
05-23-2010, 01:31 PM
I haven't noticed anyone mentioning using a rack gear form to generate the involute.

This is probably the millionth time around for this topic but in the past iterations racks have been mentioned as well as using shapers to implement the idea.

dp
05-23-2010, 01:33 PM
What I am interested in seeing is how this "button cutter" is going to cut the spline teeth. I don't see how it will work and am looking forward to seeing it in action.

The button method creates the cutter, not the spline. Once created the cutter has teeth milled into it and is relieved. It looks like an off-the-shelf gear cutter when done.

From Ivan Law's book:

http://metalworkingathome.com/images/buttonmethod_sm.jpg

John Stevenson
05-23-2010, 06:13 PM
I'm not bothered about posts getting in the way, it proves you are not talking to an audience of one :D

This pic off my 'site'

http://homepage.ntlworld.com/stevenson.engineers/lsteve/files/30_32_35.BMP

Is the one I think was mentioned. There are 3 profiles, one for 30 teeth, 32 teeth and 35 teeth, which is which I can't remember but you can see there is a difference although it's only small.
This is why commercial form cutters can work in a band between numbers.

To be honest I would imagine that normal working tolerances put a greater error into the pot than the shape of the cutter.

The rack cutter has been mentioned before but the planing action required means it limited to a shaper or specialised machine. As most people use a lathe or mill this limits you to a rotating cutter of some form.

.

J Tiers
05-23-2010, 09:31 PM
If you are going to make a cutter, which will do only the single part, that is fine....

But if you think you might make more of slightly different related numbers of splines or teeth, it might pay to make a hob......

In the simplest description, a hob is a rack form that is used as a rotating cutter..... You can, if the blank is gashed first, even let the hob drive the work, although that sounds dicey and I have certainly never tried it.

And it makes ANY number of teeth of the same pitch and pressure angle.

Aside from the relieving, which you would presumably have either way, a hob is not a lot different from a form cutter in terms of trouble taken to make it. You don't need to make a tool before you can make the hob, at worst you might grind a cutter.

John Stevenson
05-24-2010, 04:04 AM
Jerry,
As you mentioned, but quickly passed over, relieving.
Relieving on a single cutter is quite easy, a simple jig or holder, relieving on a hob isn't so simple.

The only reliable way I have seen is by a guy called Charles Lessig who has modified a Eureka attachment to do just this.

http://www.youtube.com/watch?v=kJ8kyC_bpHs

Now I'm trying to write this up as simple as possible for people who lack skills, time and machinery. No one wants to build a Eureka attachment for one hob or one or two gears.

To be honest by the time you have built the attachment, made the hob, had it hardened all you finish up with is a spurious hob made from not the best materials and you might as well have bought one.

You also need the hobbing machine, I'm fortunate I have a hobbing machine, I also have a full sized [ 1" bore ] Eureka but if I was to go this route I know I'd out price the job and although possibly interesting very few if any of the readers on this forum could duplicate what I'm trying to achieve, a simple reasonably accurate cutter made with basic tooling.

John Stevenson
05-24-2010, 11:44 AM
I thought of making a different thread, but I suspect that would be qiuckly"obliviated" because it does not agree with the "expert" prejudices.
Bloody good idea Jerry.
I for one would love to see another method of hob making other than the one Charles Lessing was put up.

[EDIT ] I'm very serious here.

I was in email with Charles over this and the Eureka attachment and Charles method because of the extended maths was to fit a follower thread to the Eureka which does away with the leadscrew whilst relieving.

Believe it or not but there are different ratios employed between cutting the spiral and relieving it because the hob rotates at the CP when cutting but it's CP / number of flutes when hobbing.

So when you are ready please roll out the words and music.

.

Circlip
05-24-2010, 12:21 PM
How many shafts are you doing Sir John??

Get yer own egg, I'll supply the pin.

Regards Ian

John Stevenson
05-24-2010, 02:45 PM
Ten.

But they are now deciding whether to go a different route.

I'll still do the cutter in spare time and show the steps now that I have started and I'll hope they change their mind and still want them doing.

J Tiers
05-24-2010, 07:13 PM
Bloody good idea Jerry.
I for one would love to see another method of hob making other than the one Charles Lessing was put up.


Believe it or not but there are different ratios employed between cutting the spiral and relieving it because the hob rotates at the CP when cutting but it's CP / number of flutes when hobbing.

So when you are ready please roll out the words and music.

.

I am aware that the spiral makes the circumference (based on 360 degrees) different for a straight and a helical path....... If you start out right, you may end up different. Which is one reason for USING a spiral gashing, since you can make the relieving come out right by simply ignoring the problem and gashing to make up.....

However, what method were you proposing to use for relieving your cutter?

If you say a relieving attachment, then you cancel out the simplicity of the original cutter "generation" method. Cancel out is a misnomer.... it adds hugely to the trouble, like making a Quorn so you can sharpen your knife....

that leaves a number of other ways,...... none as good at generating a nice smooth "correct" relief.

SO DON'T.................

Any cutter worth a dang should last quite a number of parts before needing to be resharpened..... certainly 20 parts if you use appropriate materials and don't need to cut inconel etc, etc....

if you make the cutter easy enough to make, it isn't a problem if it isn't resharpenable 20 times.... particularly for the HSM who may never need thousands of gears.

Admit up front that you may resharpen ONCE, maybe, and just "relieve" by cutting away, leaving just a short "lip" to be form cut. Or form cut first, and "relieve" later....

Since you are making no particular effort, and are simply clearing material, the order of operations is just whatever works...

And since you are making a "hob" no longer than it has to be made, the number of flanks to cut is less.

Alternately, use the "John Stevenson" relieving technique (whatever that is ), and re-index every couple gashes if you need to in order to keep "in step"

Either way, you will end up with a workable cutter, with a known limited life, but which will save you several hundred $$ over buying if you need a number of gears of a given DP but very different numbers of teeth.... And you make ONE tool, needing to "relieve" only a few turns of fluting..... which generates good gears.

If you need to make several separate cutters, that is several holders, and possibly several buttons, without saving the trouble of coming up with some method of relieving.

John Stevenson
05-24-2010, 07:36 PM
Jerry,
I'm not making a hob, that's your job because you say it's so easy.

I'm making a form cutter that is applicable to a small range of gears [ splines in my case ] which the average Joe can transpose for the odd gear they may have to make in the home shop.

I fully agree that a hob is a good deal better but the making of one and the using of one i.e. a hobbing machines isn't a quick and dirty job, like making a Quorn so you can sharpen your knife....

The forming and relieving will be done in the same operation, with a jig, but that simple if you can't make it from scrapbinium in ten minutes you really ought to take up knitting.

[EDIT] And it will be resharpenable right up to the last bit of cutter.

Got held up this afternoon when I was going to do the button tool, local firm broke the drive on a large packaging machine, two plastic sprockets, one dropped a couple of teeth and the other shattered. spares are 4 weeks lead time from Germany, line is down until it's fixed, so drop everything and belt a couple out in alloy.

http://www.stevenson-engineers.co.uk/files/sprockets.jpg

.

dp
05-24-2010, 07:43 PM
I wonder if this had been a shop-made hob thread how quickly it would have turned into a shop-made form cutter advocacy thread.

FWIW, I am actually making the Eureka and I don't even have a need for it. I just want to do it because it is a good project. Some day my kids will put it on Ebay with the admonition that while it is in good condition, they have no idea what it is used for.

lazlo
05-24-2010, 09:15 PM
FWIW, I am actually making the Eureka and I don't even have a need for it. I just want to do it because it is a good project.

The Eureka is a project in this months' Model Engineering Workshop.

lazlo
05-24-2010, 09:17 PM
Got held up this afternoon when I was going to do the button tool, local firm broke the drive on a large packaging machine, two plastic sprockets, one dropped a couple of teeth and the other shattered. spares are 4 weeks lead time from Germany, line is down until it's fixed, so drop everything and belt a couple out in alloy.

http://www.stevenson-engineers.co.uk/files/sprockets.jpg

Bravo!! :)

Now tell us how you did it! How did you make the teeth and the square spline?

dp
05-24-2010, 09:31 PM
Bravo!! :)

Now tell us how you did it! How did you make the teeth and the square spline?

Looks like everything could be done with a 6mm end mill on a CNC machine, and a die filer to square the corners.

dp
05-24-2010, 09:33 PM
The Eureka is a project in this months' Model Engineering Workshop.

Aw jeez - now I gotta go look :)

boslab
05-24-2010, 09:45 PM
Must be great when you can just about make any damn thing, and if you cant make it then you can make a thing to make the thing, i spend my time just fixing up ill machines using off the shelf bits and a bit of basic hand/machine work, i have a lot to learn i think, i must add CNC to the ever growing list of things to try, well done John
mark

dp
05-24-2010, 10:19 PM
Charles Lessig built one then added a pattern screw to it so he could relieve hobs. http://www.youtube.com/watch?v=kJ8kyC_bpHs

His thread is here: http://bbs.homeshopmachinist.net/showthread.php?t=30041

In that thread Lazlo has the link to the spudster's version but it's since suffered link rot.

J Tiers
05-25-2010, 12:14 AM
The OTHER method of relieving that I know of is the clockmaker method, which is to make the form tool with NO relief, and gash it.

You then slit the cutter angle-wise under each tooth, and squash the teeth down, which bends down the back of the tooth, making a relief and allegedly doesn't affect the form too much. This one is apparently as old as the hills, I saw it in the Goodrich book on watchmaker's lathes. Another method I have not tried.

NOT applicable to a hob, only to a single turn form cutter.

As for making a hob "being so bloody easy"....... since it involves (or CAN involve) the same damn operations as making a single turn cutter, either the single turn cutter is "so bloody easy" or both are "too bloody difficult".

Cutting the form is reasonably easy. It is basically similar to cutting an Acme thread.... angle per requirements, 14.5 or 20.

Gashing the flutes can be quite easy, if a straight flute hob is made.... Obviously either a slow spiral OR a straight flute would be a milling machine operation to gash. I'd not try it on a lathe.

The hard part is relieving, and that is the same basic problem whether you are making a circular form cutter, or a helical form cutter. I don't see a huge reason to classify one as impossible while the other is so easy as to require taking up knitting instead if you can't do it....

Maybe you can show me a good reason, but it seems that if you can relieve one you can relieve the other.

Tiffies method does NOT apply to any round cutter, it applies ONLY to a single tooth cutter such as illustrated in tiffie's picture..... that one is also in Goodrich, as is a variant or two of it, using different diameters instead of an offset.

IMO gear cutting already takes a long time, and having one tooth instead of ten makes it take ten times as long....and need sharpened after doing one-tenth the work. But I will happily agree that it can be done, and IS done all the time.

If I am going to go to that much hassle, I want a useful tool out of the time..... but that single tooth cutter gets you working for one part quickly.

The slit and squashed tooth method makes a round multi-tooth cutter more-or-less equal to any other gear cutter. I think it is a bit of a kludge, but it is claimed to work.

Now if you have another way of relieving a round form cutter as easily as that, we're all ears............

dp
05-25-2010, 12:35 AM
The OTHER method of relieving that I know of is the clockmaker method, which is to make the form tool with NO relief, and gash it.

I've used a similar method to make hooks from tubing. No reason it can't work to some extent.

LES A W HARRIS
05-25-2010, 03:58 AM
John,

don't have B.S. for splines, presume it will be close to A.N.S.I.

Here is data for class 5 ANSI Flat Root Side Fit.

Data:
Involute points are at columns "K & L" under the triangles. Used 3 point arcs thru OD/PD/& Root dia's for drawing.

http://i37.photobucket.com/albums/e97/CURVIC9/2010%20GEARS/QuattroPro-CDocumentsandSettings-1.jpg



Drawing:
slightly smaller arc, spread & infeed different.
Included M.O.W.'s 0.0840" standard.
http://i37.photobucket.com/albums/e97/CURVIC9/2010%20GEARS/VisualCADD403-12T20_40DP30PAEX52520.jpg



Cheers,

John Stevenson
05-25-2010, 04:08 AM
Looks like everything could be done with a 6mm end mill on a CNC machine, and a die filer to square the corners.
Very close.

About 30 minutes to go from damaged sprocket to CAD drawing, then CAM in Dolphin but the drawing has no resemblance to a finished sprocket, just outer shape and inner square.

Diameters widths etc are just hand written on, this is what I said in an earlier post on CAD drawing, the customer doesn't want a drawing so why waste time doing pretty models.

There is a chuck usually mounted on one end of the CNC bed so turned blank in the chuck, centre up and let rip with a 12mm end mill, only chosen because it was new and could go 42mm deep ;)
That did the inner square and the outer profile, swap to 8mm cutter to do a clean up round the outside, don't bother with the square because it will always have radiused corners anyway.

Swap to slotter and trim the corners.

One hour to turn two blanks, none of the sizes are real critical and the OD is oversize anyway for the CNC to finish.

CNC time 42 minutes not allowing setup. [ edit ] Per part that is

Slotter about an hour for both as setup was done whilst the CNC was working and same for the first sprocket.

Plus about an hour to get all the chips from out my shirt.

Just got a phone call this morning, they are that pleased they want 4 as spares.

.

Circlip
05-25-2010, 07:15 AM
Plus about an hour to get all the chips from out my shirt.



Hope yon safety bird in't reading this John.:D

Tony Ennis
05-25-2010, 07:54 AM
Must be great when you can just about make any damn thing, and if you cant make it then you can make a thing to make the thing,

...and get paid for it.

dp
05-25-2010, 12:55 PM
That looks like an Adcock & Shipley pedestal grinder with the Aussie multi-tool belt attachment. What's that got to do with anything?

In John's photo of the new sprockets a grinder is seen in the background. It's a general purpose bench grinder. Not sure what it would be the subject of comment - we all have them.

John Stevenson
05-25-2010, 02:12 PM
That 'grinder' is a de-burrer fitted with Scotchbrite wheels.

Not very good for sharpening tools :D

Mind you I've saved this picture

http://i200.photobucket.com/albums/a...lt_sander2.jpg (http://i200.photobucket.com/albums/aa294/oldtiffie/Belt_sander/Belt_sander2.jpg)

It's the only thing of tiffies that's not posed and been used - once. Like that WWII Italian rifle, never fired only dropped once...

John Stevenson
05-25-2010, 06:39 PM
John,

don't have B.S. for splines, presume it will be close to A.N.S.I.

Here is data for class 5 ANSI Flat Root Side Fit.

Data:
Involute points are at columns "K & L" under the triangles. Used 3 point arcs thru OD/PD/& Root dia's for drawing.

http://i37.photobucket.com/albums/e97/CURVIC9/2010%20GEARS/QuattroPro-CDocumentsandSettings-1.jpg



Drawing:
slightly smaller arc, spread & infeed different.
Included M.O.W.'s 0.0840" standard.
http://i37.photobucket.com/albums/e97/CURVIC9/2010%20GEARS/VisualCADD403-12T20_40DP30PAEX52520.jpg



Cheers,

Les,
Interesting that you have different values for the buttons and the centre distance.
You have 0.259 diameter and 0.289 between.

I drew this out and superimposed it on my drawing and it's close but not as close as the values I came up with.

That's to be expected as my values were derived from the drawing but is my drawing correct ? I only have the program to go on.

In any case the errors are quite small and I dare say both values will a fit a real world application.

LES A W HARRIS
05-25-2010, 09:00 PM
Les,
Interesting that you have different values for the buttons and the centre distance.
You have 0.259 diameter and 0.289 between.

I drew this out and superimposed it on my drawing and it's close but not as close as the values I came up with.

That's to be expected as my values were derived from the drawing but is my drawing correct ? I only have the program to go on.

In any case the errors are quite small and I dare say both values will a fit a real world application.
As I stated, I have no BSI spline specs to go on, also does not the button method figure the radius for the curve from the pitch dia & tangent at the base dia with nominal CTT, [Pi/40] ?

Just triged it out,it does not appear so.


Just curious, go make chips, you still have to earn your living,


Cheers,

John Stevenson
06-07-2010, 08:15 AM
Now if you have another way of relieving a round form cutter as easily as that, we're all ears............

Keep listening, the job has now got clearance and they need 10 for Friday for a job at a power station in Spain.

We are still awaiting your hob post BTW..........

.

John Stevenson
06-07-2010, 10:28 AM
Yer Hoo we have a winner, 3:15 pm here, started at 10:30.

Buttons made and holder, buttons hardened and ground.

Blank cutter made.

Jig made out of scrapbinium [TM]

Cutter relieved and flutes cut and hardened.

Test shaft made, fits across wires.

Film at 9:00, well nine my time................

Gotta go turn some blanks up.

.

Circlip
06-07-2010, 02:25 PM
Clever sod :D Make sure they pay you double. And then add 3000%

Regards Ian.

John Stevenson
06-07-2010, 04:38 PM
OK so on page one we got to the pretty picture of the spline and how the buttons are worked out which then went on the the button cutter.
This is the button cutter being prepared to take the buttons, nothing fancy just a piece of scrapbinium [ TM ] tilted in the vise to about 5- 7 degrees to give front clearance and two holes drill at 0.300" centres from the original drawing.

http://www.stevenson-engineers.co.uk/files/button%20tool1.jpg

Flat put on the top so the buttons, made out of silver steel [ drill rod ] can sit flat, one button fitted and one lying loose.

Button tool finished. Being a dedicated and trained rough arse I welded the buttons in, then hardened them and surface ground the tops.

http://www.stevenson-engineers.co.uk/files/button%20tool2.jpg

Now onto the cutter.

This is a piece of 40mm [ 1 1/2" roughly ] silver steel that I had, slice cut off and faced up both sides and drilled 8mm to suit an arbor I have for the POS Bridgeport.

Now to mark it out, line scribed on where I want the depth of the teeth to be in the lathe and using a MKI dividing tool, in this case a brken motor shaft stuck under one jaw I scribed a line then repeated it for all three jaws.

http://www.stevenson-engineers.co.uk/files/splinecutter1.jpg

These marks were drilled 4mm to take a small cap screw and next was the jig for relieving.

Someone give Jerry a nudge and wake him up in case he misses this next bit.

The jig is just an off cut that can be held in the chuck and had a tapped hole for a retaining screw, 8mm in my case and a 4mm tapped hole for the locating screw.
The important bit is opposite this locating screw is a flat 1/16" deep on the part that goes in the chuck.

http://www.stevenson-engineers.co.uk/files/splinecutter2.jpg

The reason for this is very important and it's because I'm idle and can't be bothered fitting the 4 jaw to offset the jig hence three flutes on the cutter.

When it's fitted it will run off centre and the arc it describes will make the relief.

End of this post due to the 4 pic limit, hang on for the next post so we get continuation - thanks.

.

John Stevenson
06-07-2010, 04:53 PM
Once fitted into the chuck with the flat on one jaw the whole rats nest throws off by 1/16" A light facing cut is taken so it cleans up in an arc from one hole to the next. Then the cutter is rotated to the next hole and the same operation, to the same depth is done on that 1/3rd arc, and again on the last.

http://www.stevenson-engineers.co.uk/files/splinecutter3.jpg

So we finish up with 3 egg shaped arcs that meet at each hole, as in the pic above.

The light will be coming on for some people, JERRY, you still awake ?

Now the same process is done but using the button cutter on each arc to the depth given in the first drawing.

http://www.stevenson-engineers.co.uk/files/splinecutter4.jpg

Again you can see the transition mark above. Once all three arcs have been done the cutter can be removed and three cutouts made.

The cutting edges have to be radial lines that clear the locating hole and disappear sideways to give room for the chips and to remove the part of the arc that is increasing back to major diameter.

http://www.stevenson-engineers.co.uk/files/splinecutter5.jpg

And the finished cutter after hardening but before grinding, sorry about the crap shot.

http://www.stevenson-engineers.co.uk/files/splinecutter6.jpg

Provided you always grind all three faced radially and equal this cutter will have relief until it completely worn down.

Again pic limit.

.

John Stevenson
06-07-2010, 05:01 PM
Last shot that wouldn't fit in.

http://www.stevenson-engineers.co.uk/files/splinecutter7.jpg

Cutter on the POS and test bar in the dividing head, 12 splines cut in two passes at 150 rpm to save wear on the cutter. I actually made two whilst doing this just in case.

The idea isn't new, Unwin used a similar method but he spent more time and used more teeth, this way everything follows on from the 3 jaw chuck. No drawings were used other than the original layout drawing and only three microns were harmed in the making of this cutter but I don't know what happened to them.

And yes I know it will only do a small range of teeth in one size unlike a hob but it's aimed at the simple one off that someone may want to do in the home shop.

.

carlquib
06-07-2010, 05:58 PM
Brilliant, that three tooth bit. The ones that I have made only have a single tooth with the hole offset a bit to provide clearance. I will have to give this cutter a go the next time the need arises, ought to be three times as fast.

Deja Vu
06-07-2010, 06:33 PM
Yes, Brilliant! But i'm not impressed. I've grown to expect as much from you, JS. I suppose you kept your local pub busy for a spell after completion? :D

John Stevenson
06-07-2010, 07:10 PM
No I got dragged screaming and kicking down to Asda [ Walmart in the US ]

Still managed to sneak two packets of fig rolls and a big bag of pork scratchings in the trolley.................

Pub tomorrow night, and Wednesday and possibly Thursday if my new breakout boards are ready.

oldtiffie
06-07-2010, 08:38 PM
Nicely done all round John.

That should well and truly skewer a few Doubting Thomases.

J Tiers
06-07-2010, 10:28 PM
I don't "doubt" a damn thing, Ralph. So I'm safe from "skewering".......

It's a good idea... make fewer edges, and you need fewer reliefs..... compromise between hassle and speed of cut, and there you are. It's an extension of one of the ideas in the Goodrich book.

One would probably top out at about 5 edges doing this, more is a hassle, might need a smaller button cutter, and likely has no real advantage Three is a good compromise.

lazlo
06-08-2010, 12:35 AM
Beautiful job John! That's the first time I've seen an amateur cut an involute spline. But then again, you're hardly an amateur :D


And the finished cutter after hardening but before grinding, sorry about the crap shot.

Question: why not cut-out the relief slots before hardening?

willmac
06-08-2010, 02:41 PM
"If the width of John's cutter was equal to the circular pitch of the gear/spline wide it might have been one of a set of ganged cutters mounted on and keyed to a common arbor"

I don't think that even with ganged cutters made by John's method it could be considered as hobbing. The shape of the 'hob' would not be correct.

What John has done is show a really clever (simple) way of creating a relieved form cutter with good enough accuarcy for most of us. Definitely one to keep.

John Stevenson
06-08-2010, 04:36 PM
I don't think that even with ganged cutters made by John's method it could be considered as hobbing. The shape of the 'hob' would not be correct.


Correct.

A hob has straight sides that generate the curved involute, my cutter already has the curve generated by the buttons. Ganging some up would do nothing at all.

The offset method isn't new, as I said earlier Don Unwin used a similar method that was in ME in the 60's, 'my' method just takes a few steps out to make it easier.

Lazlo, the relief slot were cut before hardening it was the final grinding I was mentioning, just a lick on all three faces freehand after hardening.

Anyway, all 10 shafts done and dusted.

http://www.stevenson-engineers.co.uk/files/splinedshaft1.jpg

And will be on their way to Spain tomorrow.

Got to be one of the hardest seven pounds eighteen shillings and seven pence I've earn't this month so far.

The cutter stud up to the job remarkably well and other then the initial quick lick on the grinder it did all 10 with no problems.

http://www.stevenson-engineers.co.uk/files/splinedshaft2.jpg


Just thought this would help for any quick and dirty gear cutting jobs that may arise where you don't want / need / afford [ delete as necessary ] to get a cutter for a one off or in my case a cutter that is next to impossible to buy.

.

oldtiffie
06-08-2010, 05:27 PM
Originally Posted by willmac

I don't think that even with ganged cutters made by John's method it could be considered as hobbing. The shape of the 'hob' would not be correct.

I sure got that 200+% wrong and should have known better - I did - but I would insist on bull-$hitting and not only painted myself in a tight corner but also dug myself into a bloody great hole - and kept digging.

The "hob" would not only have had to have straight sides but would need to have been spiral cut as for a screw thread but with a transitional circular pitch equal to the circular pitch of John's job and cut on a relieving lathe (Eureka device??) with a pitch modified by the Cosine of the "thread" helix angle etc. The cutter in use would have needed a universal mill tilted for the cutter helix angle and with a universal dividing head geared/connected to the cutter arbor etc.

My complete cock-up.

Silly me.

My apologies - as I don't want to distract further from John's excellent work.

oldtiffie
06-08-2010, 05:36 PM
John.

Congratulations on a really really very nice job and demonstration of Ivan Laws book (amongst others) "button" method of making a cutter to make a gear or a spline and then showing that it works.

As a matter of interest, did you have a chance to use the mating part ("female") to test for fit or did you just have to go ahead anyway?

I was interested to see just how well it went as regards "fit".

I hope you don't get paid with a Spanish Bank cheque as their banks got themselves into a bigger hole than I dug myself into.

leesr
06-13-2010, 06:59 PM
JS

Nicely done , there is always an other way to skin a cat.
did you have a mating part to check to? & did you verify over wires?

reiterate liked you post very much. keep up the good work.
very impressive.

Cheers

Lee Miramontes

John Stevenson
06-13-2010, 07:11 PM
No mating part, I asked for one but it appears it weights 80 tonnes and is in Spain so took a measurement over two teeth, side to side from the pattern.

These are not a super critical part, they have a 3/8" drive welded on one end and are used to wind a part out for inspection, they are not a drive shaft.

leesr
06-13-2010, 07:18 PM
Just wanted to add that a hob is basic worm it can be Threaded just like a thread. & if it has more the one start it can be calculated like a helical gear, if it a single start then it has to be calculated as a worm. just other options:D

By single pointing or by thread grinding, what needs to be calculated is the
lead angle, thus then the "helix angle" 90 deg - lead angle = helix angle, & the pitch, and so on. One way the hob can be measured is by measurement over 3 wires or 2 wires depending on the calculation. the same gear wire can be used.
any way for mfg this one way it can be done.

Cheers:D

Lee Miramontes

leesr
06-13-2010, 07:39 PM
No mating part, I asked for one but it appears it weights 80 tonnes and is in Spain so took a measurement over two teeth, side to side from the pattern.

These are not a super critical part, they have a 3/8" drive welded on one end and are used to wind a part out for inspection, they are not a drive shaft.


JS

Gear shops will use spline gages for a go & no go if cost permits , or a sample spline cut internal" in this case", there is a way of calculation a 2 wire check "since a spline has a truncated adddendum, " on gears a 3 wire check can be accomplished. what this does is tell you if the spline involute is correct as cut.

spline gages are functional part which will allows for all errors on a spline.
for example, lead (parrallel error), involute error, & index error.

depending on the # spline teeth, DP & PA the index, lead , & involute will very.

a good general for Hobby work these errors will work .0015" max index error
.0005" max lead error. involute .001" max, or better yet if it fits the mating part:D

Keep up the Good work JS & thanks for sharing your information.

Cheers
Lee Miramontes

late night fubar my postm, re-edit to fix my poor writting. senior moment.

j king
06-13-2010, 08:55 PM
Ill add my compliment, Nice job John.

Ozzie34231
11-06-2010, 11:07 AM
Luv it John!
Just got here a few days back and to Gearotica as well.
I posted there, that maybe Art could have his program generate G-code for the lathe to cut the exact tooth form for a cutter like yours. That would eliminate the whole button making process.
Got to get my hands on some 1 1/2" drill rod before I bust with the need to make a cutter. I'm thinking I can make a dovetail tool using your method; just need to dish down the center for mounting.
Did you draw down the hardness; how did you control it; wife's oven?

Thanks for sharing.
Ozzie