Wire EDM gear manufacture 1

[nsmith40]

I have looked for information relating to wire EDM as a method of gear manufacture but have not been able to find too much useful information.

Does anyone have any experience in using wire EDM for cutting steel gears? What kind of surface finish is to be expected?

 

[MikeHalloran]

My machinist friend got a specialized calculator with his (plunge) EDM machine that predicted cycle time and finish, among other things. The best he could do was a really smooth matte finish that measured better than it looked.

You might ask about the fatigue life of an EDM surface before you get real excited about making gears that way.


Mike Halloran
Pembroke Pines, FL, USA

 

[israelkk]

EDM of gear is very expensive compared to hobbing and less accurate. However, for quick prototypes it may be OK. NASA tried to EDM machine 65RC steels for space use in dry environment with some success and manage to get AGMA 10 accuracy. It was justified due to the high hardness of the gears.

One real problem is that the process produces hydrogen and that leads to hydrogen embrittlement. We has a machined helical spring that shuttered to tiny pieces when loaded after EDM machining. The spring was EDM machined after heat treatment. To solve the problem the spring was EDM machined and them heat treated.

 

[nsmith40]

This is indeed for a prototype, so the cost of EDM is not of great concern.

Thanks for the info on hydrogen embrittlement; I was trying to determine if heat treat would be better before or after cutting, but it seems that heat treating after cutting is the way to go.

Anyone have any more input on wire EDM as a gear manufacturing technique?

 

[tripleZ]

Your length will be of particular importance, as will the age of the wire EDM machine. Newer machines have done a very good job with compensating wire guides and tension monitors to eliminate "barreling" or "hour-glassing" in longer gear forms. Older machines are much trickier to manage on the longer parts.

For your cycle, you will need to take a rougher pass often followed by several finish passes to get the proper size and surface finish. Newer machines will leave you with a .0002"-.0004" per side recast layer that is normally polished off to get down to good base material. Older machines, again dependent on part length, might require .0008" per side of stock removal to fully clean up.

How long is your gear? Is it internal or external? And what is the profile tolerance and surface finish of the form? The answers to those will give you a good starting point.

 

[gearmold]

The EDM in "wire EDM" stands for Electric Discharge Machining. This means that metal is being removed from the workpiece electrically, or more specifically through millions of tiny sparks which jump between the wire (from the first part of the name) and the workpiece. The process creates extremely high temperatures which melt a little pit at both ends of the spark. This means that some material is blasted out of both the workpiece and the wire with each spark. We don't really care too much about the wire, because it is constantly moving, and renewing itself, but the workpiece is eroded away for a certain distance around the wire (the spark "gap").

As the workpiece is eroded, the table moves (X-Y) to maintain the flow of electric current. A computer in the machine interprets the user program to determine X and Y table movements, and hence, the 2-dimensional shape. Yes, it is possible to cut a limited 3-dimensional workpiece, but I don't believe that is your intention. By the way, the X-Y table movements on modern machines are in increments of 10 microinches or less, so I must question the idea that hobbing is more accurate than wire EDM. Ground gearing can, however, be more "accurate".

The surface problem with wire EDM arises from the fact that workpiece material is molten when it is blasted away from the surface. Most of it is flushed away by the dielectric fluid surrounding the cut. At the extremely high plasma temperatures of the spark, however, many things happen, including turning the dielectric fluid into a gas and sometimes blasting the molten material back onto the workpiece surface in a different location. When molten material is redeposited onto the workpiece surface, it is called "recast".

This recast consists of material which has, literally, been cast back onto the surface, so it is very hard. Underneath the recast layer (and sometimes considered a part of it) is a zone of material which has been heated and annealed, so it is softer. A hard, rough outer crust with a softer layer directly under it, does not generally make for a good working surface.

People who run wire EDM machines generally want to cut as quickly as possible, so they can get the next job into the machine to help make their monthly payements. The higher the spark settings, the faster the machine will cut. This is only limited by the fact that material is also being blasted away from the wire, so too high a spark current/time setting will cause the wire to break.

High current/time settings result in faster cutting speeds, but also in deeper pits, a rougher surface finish, and a thicker recast layer. That is why, as was previously mentioned, most modern machines do a primary, rough cut, and then several finish cuts at lower settings. These finish cuts bring the workpiece to size, but also smooth the finish and minimize the recast layer.

There are several problems associated with wire cutting a gear which is intended to operate, rather than a cavity used for molded parts. First is the problem of center location. The center hole can be wire cut before the gear, and the table-positioning accuracy, as I have mentioned, is excellent. Unfortunately, the wire must be broken, then relocated in the wire guides of the machine before cutting the actual teeth. This operation can cause runnout error which may be extreme for your application.

The second problem arises from the fact that you are cutting a spur gear and intend to use the male portion of the cut. If the workpiece is held at the edges, after the first rough cut it will either fall or be held tenuously in place at one point on the OD. Yes, it is possible to hold this more securely by leaving it attached in several places, but then we go back to the first issue... we have to break and relocate the wire even more times!

Finally, remember that steel expands and warps through the heat treating process. The wire EDM process doesn't "care" about the hardness of the material. Only the conductivity and melting point matter. If heat treating is done after the wire EDM process, the accuracy of the process itself is overshadowed by the expansion and warpage of the heat treat process.

Basically, the wire EDM process is not a really good choice for producing accurate operating gears, but an excellent choice for producing a female gear shape, such as used in injection molding cavities.

 

[jmaurel]

A friend of mine have created (and sells)a software for this purpose, It is efficient for helical internal gears as the tool price is important and the helix angle small.

 

[plasgears]

My experience is wire EDM of PM tools and production by PM methods. Wire EDM is a tool maker's approach.

 

[Vdot]

I have produced a number of gears using wire EDM, and can assure you that it is a viable means for producing very accurate, specialized spur gears. I am certain that class 13 and higher is readily achievable. As to surface finish, I have produced 10 microinch finishes without any problem. There are some interesting comments in the thread regarding the suitability of wire EDM for gear manufacture. I am most interested to learn if this is speculation or experience. Fortunately, I don't have to speculate. Two gearboxes I produced that employ wirecut gears are successfully transmitting 2800 hp each with input speeds in excess of 15,000 rpm. Incidentally, these are carburized, high contact ratio gears made of 9310. Most other details are proprietary. I am certain that there is great opportunity to utilize wirecutting to produce high-performance gears.

 

[israelkk]

Vdot

After carburizing the gears, was there any other procedure such as grinding. etc? I assume the gears were cut in the normallize condition before heat treatment.

 

[MikeHalloran]

Please educate all of our membership by posting photos of the finished gears.



Mike Halloran
Pembroke Pines, FL, USA

 

[Vdot]

The gears were wirecut before and after carburizing. I will post some pictures of the gears as well as the application on our company website fairly soon. The fact that these were the first gears that that we had ever manufactured (I literally had never made a gear previously; we own no conventional gear-making equipment), suggests that this is a very robust method. Thank you for your interest.

 

[israelkk]

Vdot

If the gears were cut after carburizing (in the high strength condition) it is a risky approach because there is a good chance of hydrogen embrittlement cracking of the carburized layer. Future batches may develop cracks of the carburized layer even if the first batches are OK. It is a risky approach to wire EDM the teeth after carburizing. I would suggest nondestructive test such as magnaflux of the gear teeth for cracks. If this is for aerospace use I would check with the metallurgy department and the codes to see if it is applicable at all.

 

[Spurs]

nsmith40

I have produced and used wire edm gears in many prototypes that I have produced for clients with great success. These prototypes where however not life cycle tested for for fatigue life as I would never use any prototype gears produced in one method to predict life cycle for a different material and process.

With care and extra passes, the surface finish can be excellent although admittedly never be as good as ground gears. I have seen the surface finish better than hobbed gears and even shaved gears.

The decision of what material to use is quite important for wire edm as it is with any other gear manufacturing process. Issues to consider are not only the application requirements, but also the stability of the material during the wire edm process. You dont want to use a material that will spring back as it is being cut.

The accuracy comments made by others in this thread is not consistent with my experience. Not only is wire edm more accuract than hobbing, and shaving, but it has many features than can even be better controlled than grinding. Master gear qualities can be achieved with the wire edm process.

I would be happy to look at your specific applications and help you to understand if wire edm is in line with what you need. I could also provide you with finished wire edm parts.

See website for more information.

http://www.webgearservices.com

 

[Pressed]

A few things to keep in mind:

Not all gears are created/utilized equally.

Not all EDM machines can produce the same quality surface finish. Older machines are still used and can be the sole source of one's experience.

As I mentioned in a different thread, there are CNC bamboo scrapers used after EDM'ing in Japan; surely not a common practice in other countries/shops.

And for the weary, I take no sides.


==========================================
Business Site

http://mech.e.tripod.com

------------------------------------------
Cycle Utopia.......

http://www.tailofthedragon.com