|Previous article||Next article|
Tips for JIT Gear Manufacturing and the Job Shop (Continued)
In the last article, we discussed the real possibility that the right manual equipment in the right gear shop with the right software can compete for JIT and repair work. Now we need to look at some of the situations on the shop floor that can challenge both the beginning and seasoned "Gear Man". Most of what you find here is based on "Rule of Thumb" suggestions derived from years of experience. If anyone cares to elaborate or challenge anything said here please
Tool inspection involves detecting wear and applying proper sharpening. Don't attempt to sharpen gear tooling without the right equipment and know-how. It's not brain surgery, but it is easy to make costly mistakes.
There are grinders made especially for shaper cutters, and other grinders made for hob cutters. These don't look or work like surface or cylindrical grinders you may be used to.
If you don't have tool sharpening in-house, a local source can be very important. Otherwise you will need to maintain an inventory of spare tooling to cover times when cutters are shipped out for service.
Usually a small job will finish without the need to resharpen. This is because operations like shaping and bevel gear cutting make full use of all cutting surfaces. Hobbing is different because the hob must be shifted on its axis to distribute wear. To maximize the number of parts between hob cutter servicing, the cutter must be shifted on schedule - say, 10 loads per hob position. The idea is to distribute wear over the full length of the cutter. Some hobbing machines automatically shift to distribute wear. Otherwise the operator will need to manually shift the cutter.
Going too long between sharpenings can quickly damage any type of cutter and require excessive stock removal. This kind of damage can be especially fast when cutting harder materials. Always inspect a cutter at removal to determine wear. The rule of thumb is that you do not want to see light reflecting from cutting edges. Use magnification to see radius, wear, or chipping. Some materials (mild steels and fine pitches) are more forgiving of cutter condition but proceed with caution when going beyond the reflecting light stage.Some materials like dead sharp cutters, particularly nylon or delrin materials and brass or bronze materials. These materials may not damage a cutter, but gear quality can suffer. Burs forming on nylon and brass get thicker when the cutter is dull, so if the bur gets more difficult to remove, then the cutter is beginning to dull. Usually these materials do better with a slightly greater tool rake angle.
Material and Heat Treatment
Material selection and heat treating are almost always specified on prints, but when you are working from samples it's good to have a backup plan. I think shops everywhere have their own recipe for material and heat treatment. A Rockwell tester can be helpful here. Here are some simple guidelines that can work well in most situations. If the gear is over 5 inch diameter and hardened, consider high carbon steels and induction or flame hardening teeth only (40-45 Rc). If you have a hardness tester, check the sample's tooth surface away from wear areas to see if it is above 40 Rc. Check again near mid-face. If there is a big reduction in reading, you can assume a high carbon steel and flame or induction heat treatment.
If under 5 inch consider a cold-drawn low carbon steel and furnace heat treatment (.010 to .030 surface penetration and 60-62 Rc) Test the sample at tooth and at face away from tooth. If hardness is 58 Rc or above, it is likely low carbon and case hardened.
For general gear service, one of the best overall materials for machining, cutting, and furnace treatment is C11L17. This is a cold rolled leaded steel with high machinability, good chip formation, and good core strength. If highest core strength in a furnace hardened gear is needed, the usually more expensive 86L20 is an excellent choice and still retains reasonable machinability and good chip formation.
Mills are trying to find replacements for Lead in steel alloys. Calcium (and perhaps other alloying elements) are used to improve machinability without lead.
12L14 is known for its machinablility, but it is a poor choice for gears. NEVER use 12L14 for a heat treated gear.
Can you produce or obtain accurate, stiff work holding fixtures?
Get up and take a break, because you need to return to this with your full attention. The first mistake commonly made is to cobble something together to mount the work. Also, fixturing often fails to get the attention it deserves when quoting the job. This is certainly an area where experience really helps.
As you know, if the work shifts in any chip making operation, you can count on breaking the cutting tool. On a lathe or milling machine this might be a $5.00 insert or endmill, but in gearing the cutting tool is considerably more expensive and you may not have spares.
If the workpiece is allowed to vibrate, you can count on poor finish and possible tool damage. The rule of thumb for minimum vibration is "thicker and shorter is better". Unless you are cutting pinion shafts, think massive, stiff and short for backing support. Any spoked or thin-sectioned gear will need support near the teeth. Avoid large thin walled support if possible - these structures often ring like a bell. Sometimes you can experiment with damping material, but this can be trial and a lot of error which can drive your costs to places you didn't plan to go.
For a long shaft with gear in the middle, use a stationary collet/bushing support close to the gear or a stationary "V" bumper mounted opposite the hob cutter close to the gear.
Whether it is one time or repeat work, CNC or manual, you need fixturing that does three things well:
a. Stiff, rigid support and non-slip drive for the workpiece.
b. accurate centering and alignment of workpiece with respect to pitch line cylinder runout.
c. If bevel gearing, provide accurate back location of workpiece. This is crucial to backlash.
The subject of fixturing can't be adequately discussed in a simple article like this. Just understand that some jobs will present unique challenges. Speeds, feeds, and direction of cut in combination with material type and coolant can either help or hurt, so be prepared to change something when the job starts. Your gear guy's experience can be crucial here.
Shops that have a wealth of fixtures and work holding methods probably don't see what the big deal is. For most simple hobbing, shaping, or thread mill fixturing there are multiple approaches that are simple and effective. Just don't be surprised if your favorite approach fails in some respect on some unusual workpiece.
How do you plan to setup your machine going from job to job?
Don't underestimate what this can require, especially with bevel gear equipment. Setting up shapers is a trival operation, gear hobbers are relatively simple, but bevel gear manufacturing requires a sequence of math calculations collectively known as a "method". Methods for spiral bevel gear generators typically require hundreds of math calculations. A complete solution (known as a "summary") is required for each gear set and these summaries when done manually are error prone and very time consuming.
The practical approach to setups and summaries is to use your PC to do the math calculations. This avoids error prone, mind numbing calculator work. Believe me, shops either use the PC or wish they could. So you have to ask yourself: do you want to program your PC to create setups or summaries, or do you want to buy the PC software and concentrate on cutting gears?
Well, the temptation is too great. I'm going to repeat the earlier plug for the software that Ash Gear and Supply sells.
(set sales talk on)
You can get fast accurate software that incorporates on-line help, "What-if" choices of tooling, and a host of other features. Downloading a demo is the best way to satisfy yourself before you buy. The programs are professionally developed, field tested and maintained through free updates, all for the one-time price of a license. This software is available for gear hobbers, thread mills, and a growing list of Gleason® spur and spiral bevel gear generators. Visit Ash Gear and Supply to read all about the features on their Software Page. Or you can visit this website's download page to get free demos for your personal evaluation.
(set sales talk off)
Inspection of finished part
Don't forget this important area - Gear inspection. This can be as simple as measurement over wires, or as high tech as inspection using CMM.
But lets back up and take a look at quality control and why it's important to you.
Quality control is a mindset as much as a program for your shop. Basically it requires that each operation be monitored to catch and correct errors before creating bad parts or passing along bad parts to the next operation. If quality is monitored at each operation, then final inspection should be a formality and a confirmation, not a cue for finger pointing and the rework police.
Investigate ways to monitor quality, such as Statistical Process Control (SPC). SPC is one of the most effective ways to determine if a machine and its tooling, fixtures etc. are actually capable of meeting the accuracy requirements you expect from its operation. It is often used to monitor machine condition and to detect when a machine and its operation begin to deviate from acceptable performance. Sample measurements are recorded over time to establish the mean value and the standard deviation about the mean. If the deviation grows over time, this usually indicates wear in the machine meaning it can't hold the same accuracy now that it used to. As long as 3 standard deviations in data collected fall well within acceptable part tolerance, the process is given a clean bill of health. If part tolerance falls within 3 standard deviations then maybe you need to start thinking about repairing the machine. Statistically, you can expect 99.7% of parts to be in spec if the +/-3 standard deviation points for the measurements collected match the +/- tolerance required. You must decide if 1 bad part out of 100 or so parts is acceptable. If you are doing close inspection on a repair job and nursing it through this machine, you're probably okay. If you are doing random sampling while producing large quantities to a customer, likely it's not okay.
This is a simplification of SPC and its application, but I hope you get the idea. Maybe someone would like to do a future article on SPC and its practical application to manufacturing?
Find ways to protect quality. For example, create a part storage system so that parts cannot come into contact. Pick out the areas where improvement can make the biggest diference. Get people together to get ideas where these areas are and prioritze.
Control quality through all the steps. Use process sheets that define the next process or station so that steps aren't omitted. Have each operation signed off by the operator and make it his responsibility to see that the job moves physically to the next station. Jobs can be forgotten if no one is responsible for moving it on. And there's hardly anything worse than a heat treated gear with a missing keyway.
Process sheets are great places to note what went wrong. Study the sheets for things that went wrong, looking for patterns or repeats of the same mistake. Decide on a policy change or procedure to eliminate future repeats. Show this plan to employees involved so that they will feel good about the change. Take the attitude that everybody wants to do a good job.
First piece inspection and sign-off by someone other than the process operator. The operator naturally looks the first part over, but you want "fresh eyes" taking a second look. This can avoid the possibility that a lot of the batch might become scrap or rework.
Institute regular inspection policy for your inspection equipment. If employees provide their own, this can be anarchy. Keep calibration and instrument inspection simple so anyone unsure of their instruments can quickly check. Be aware of temperature and the effects on close tolerance work and inspection.
You as shop owner/manager are the only person who can make quality control real in your shop. You have to to be a believer to convince others. But if you think these quality issues sound like overkill and you choose to ignore them, then it won't be long before the finger pointing and the rework take over. You'll be afraid to turn your back or go home. I guarantee it.
Gear quality as defined in AGMA standards is something else you need to be aware of. This establishes what tolerances the print-specified gear class allows you. If you need inspection equipment you can usually get help from the internet and machine tool dealers to find the best equipment for the job. Ultimately, inspection requirements will depend on your customer's quality requirements and the level of gear quality you can offer.
There are operations that improve gear quality (e.g. shaving, honing, and grinding) but these topics are beyond the scope of this article.
Perhaps this has helped a little to clear up some of the topics in gear manufacturing involved in JIT and small lot manufacturing. At the risk of repeating, buying very expensive CNC gear equipment may not be the answer for every shop. Take a new look at manual gear cutting equipment through the window of software technology. You may even see an opportunity to recycle some old yet attractively priced manual gear cutting equipment. Now that's waste reduction!
|Previous article||Next article|
If you would like to comment on this article, please your content, comments or suggestions to Meshing With Gears.
G&E, Barber Coleman, Newark, Lorenz, Lees Bradner, Pratt&Whitney, Gleason, are registered trademarks