Get Better Before You Get Bigger

This principle is probably universal: For a job shop to thrive, it needs to be more than a job shop to at least a portion of its customers.

G&G Precision in Winder, Georgia, is an example. The machining contractor doesn’t look like a job shop to its largest customer. Instead, that customer—a maker of spinal implants—sees G&G as its research and product-development partner.

One of the “G”s in “G&G” is for Rich Griffith, co-owner of the shop. He has a background in medical device R&D. By leveraging that experience for the sake of the big customer, he has helped that company refine the designs of some plates used in spinal surgery in order to make those plates easier to manufacture without compromising their function. He has also worked with the customer to develop not just the machining process, but also the design and operation of components that bring new functionality to the next generation of the customer’s spinal-plate product line.

The other “G” is for Paul Gombar Jr., the other co-owner. Formerly a machine shop supervisor, he brought the larger share of actual machining experience to the partnership. Together with Mr. Griffith, he opened this shop 3 years ago, relying on the hands-on assistance of his father, Paul Gombar Sr. An accomplished machinist, the elder Mr. Gombar came out of retirement to help, and for the first year and a half, the shop was just the three of them working together.

There were some good months of business during that time. Unfortunately, the good months never seemed to occur one after another. During the lean months following good months, there was plenty of room to doubt whether the shop would ever be a success.

“It really could have gone either way,” Mr. Griffith says.

The spinal-implant customer changed that. A friend of Mr. Griffith’s went to work for this company. By serving this customer in a way that combined machining capabilities with the shop’s additional in-house expertise, G&G created the basis for a stable relationship and a steady flow of income.

Getting this reliable income was particularly significant. Watching what the steady flow made possible for the shop taught both of the partners an important lesson about machine shop ownership. Being more than a job shop may indeed be the key to thriving. However, having at least one predictable, ongoing source of revenue is essential for a young shop simply to survive.

Here’s why:

Finding new business is inherently unproductive, the shop owners say. It requires a lot of high-value time and attention—specifically the time and attention of the owners themselves.

By contrast, recurring machining work certainly does lend itself to being performed productively. In fact, the work can be performed increasingly more productively with successive runs, as the shop learns how to perform the work in ways that require less human attention. After a time, the recurring work can provide some income that has a relatively light impact on the shop owners’ efforts and energies. If this income can just keep the shop’s doors open, then that may be enough. The owners can use the time to contact, court and connect with all of the other potential income sources that the shop needs in order to take on more staff and equipment and start growing to a healthier size.

Today, G&G has started growing in exactly this way. Its full-time employees are Cody Reed, Joey Chumbler, Don Veal and Dalton Wyatt, and the shop now has a larger range of CNC equipment than what it started with. In some part thanks to the shop’s help, the spinal-plate customer has also grown—so much that now a different challenge has appeared.

The customer’s machining needs have gone up, and soon those needs will increase even more. The result is a dilemma for G&G. If the shop can’t meet the demands, then it puts at risk a customer relationship that remains quite valuable for the shop. However, if the shop invests in new capacity to meet an anticipated level of activity from this customer, then it risks being stuck with the costs of underutilized equipment if this one big customer should change its plans, scale down or fail to increase its demand at the rate that current forecasts seem to suggest.

Certainly the new equipment will come. G&G will keep on growing. However, for now, the solution the shop has embraced is to get creative. The shop will find its next increment of increased capacity not by getting bigger but by getting better—that is, by applying ingenuity to make better use of the resources it currently has.

While those resources include equipment and personnel, the most significant underused resource for this shop is time. Every day, the shop is either lightly manned or entirely unoccupied from 6 p.m. to 6 a.m. The shop has been making use of this time, but only in the most straightforward way. That is, for parts with long cycle times, the machines keep on cutting during the night. But what about the parts with short cycle times? The spinal plates fall into this latter category.

To make these parts much better candidates for lights-out, unattended production, the shop realized it had to change its processes to more flexibly accommodate small batches. Batches needed to be set up easily for unattended runs, and a single cycle needed to be able to easily accommodate multiple part numbers with little preparation. With innovations that address these needs, the shop is becoming more automated. In fact, a basic and accessible vertical machining center will now become a reliable resource for precise, flexible, unattended machining. The shop originally purchased this machine just to drill some angled holes.

Father and son both contributed to getting the machine to this level of production. Mr. Gombar Sr. created the hardware, designing and building the indexing fixture that allows the machine to run a variable mix of spinal-plate part numbers in a single setup. Like a tombstone held horizontally, this fixture allows clamping specific to particular part numbers to be rapidly swapped in and out at each fixture position.

Now, Mr. Gombar Jr. will build upon this fixture system by addressing the programming-related challenges of getting more automated. Unattended machining is little gain for the shop if simply preparing for the unattended run commands significant time during the daylight hours. To avoid this negative tradeoff, Mr. Gombar Jr. is applying both probing and CAD/CAM technologies to allow this machining center—and other machines in the shop—to be quickly and efficiently set up for whatever mix of parts the shop happens to need to produce that night, after the formal working hours are done.

Touch And Know

The machining center that was bought for the angled machining work is a model VF-1 from Haas Automation. Haas also made the machine’s indexer. Mr. Gombar Jr. says he and Mr. Griffith did not have any demanding requirements in mind when they were shopping for this machine. They did not know they would come to rely on it as much as they are about to. Instead, he says part of what attracted him to the machine was some of its simplest features. The control unit has a tray attached with a clipboard for jotting notes, he says. The front exterior of the machine includes simple ledges to support the small plastic bins that commonly hold screws, clamps and other fixturing components. It seemed clear to him that the machine had been designed by someone who understands the day-to-day work of a machine shop.

This machine performs the more complex operation on spinal plates machined out of Titanium 6-4. Each spinal plate is a curved bracket that conforms to the backbone of a patient with spinal trauma, bridging either two or three of the soft discs that give flexibility and motion to a human spine. The plate fastens to the vertebrae adjacent to injured discs in order to hold the discs immobilized, safeguarding the patient from further damage and pain.

At least 30 spinal-plate styles are machined in the shop, and different styles have different curvatures. The curvature affects the angles of machined features, particularly the holes through which the screws fasten. Originally, the shop tried sine plates, thinking it needed a setup such as this for rigidity. The shop quickly rejected this approach because of the setup time involved. After that, it was clear that a VMC with a rotary indexer potentially offered the simplest and most cost-effective solution. With the tombstone fixture holding the parts, the indexer pivots each part to the angle needed for each machined feature.

The fixture was designed with many positions around its faces simply because there are many different spinal-plate styles. Part numbers were hard-coded to faces of the fixture. That means indexer positions were hard-coded to programs, so that each NC program would include the particular A-axis position to which the indexer had to turn in order to find that part. Only when this hard-coded value became a variable did a truly flexible process become possible. Just this change in thinking was the most important step for giving the shop new capacity.

In fact, Mr. Gombar Sr. seemed to anticipate this very thing. His fixture makes different styles interchangeable from face to face. Different part numbers could be relocated, even assigned to faces on the fly, if only the program could know where to find each part. Mr. Gombar Jr. knew how to accomplish this: by using the machining center’s spindle-mounted probe (from Renishaw) to touch each setup station and identify which part is loaded there.

Just for the probe, he will add a small bore to the clamping for every part number. Every bore will be different—a different hole size for each part. Hole sizes can therefore be matched to NC programs. As a result, the machining center’s CNC will not need to know what part it is about to produce even as the indexer is rotating to present the next part for machining. Instead, the machine will simply measure the diameter of the reference hole, and a simple algorithm written by Mr. Gombar Jr. will use this hole diameter to ensure that the correct corresponding program is loaded in the control.

Indexer rigidity has not been a problem. In the end, the process may be flexible, but the indexer is not. The shop’s initial concerns about using a fixture supported at just one end, cantilevered out from the indexer, have not borne out—there has been no deflection affecting workpiece accuracy. In fact, the setup on the indexer has proven so rigid that the shop is confident it can double the unattended machining capacity by adding another, identical fixture to the free end of the current fixture and cantilevering the work out even farther. Doing this will allow the VMC to produce the kinds of unattended machining volumes that are already being seen on the shop’s first-operation VMCs, which take advantage of a CAD/CAM feature that quickly tailors a program to precisely the batch size the shop requires.

Fixture Flexing

The angular milling and drilling on the Haas machining center represents the second operation for the spinal plates. The first operation (the first setup) machines the simpler face of the plates, where angular machining is not needed. This work is done on either of two older machining centers that the shop bought used when it opened for business.

The spinal plates are small, and the workzones of these machining centers are large by comparison. The shop can easily batch as many as 40 pieces for a single, long, unattended run, and sometimes the machines are used to run this kind of volume. The shop has its own high-density fixturing created in-house that is able to hold all of these pieces close together on the machine’s table. However, a more typical run has not been this big. It might involve just five or six pieces that need to be run right away, either day or night. Such a run leaves most of the fixture positions empty, so an NC program tailored to the 40-piece batch would cut mostly air.

Mr. Gombar Jr. says a straightforward but very effective feature of the shop’s CAD/CAM software has been vital for adapting to such batch-size fluctuations. The NC programs themselves do not have to be changed. Instead, a “multiple fixture” feature of Delcam’s FeatureCAM software allows the shop to adapt for batch-size variations of known parts with essentially no additional programming time. The programmer programs the part as if just one piece were run, then designs the layout of the multiple-piece fixturing in a separate step. The software treats this layout as a separate file, a sort of metafile that draws upon the part program as appropriate. The software can group multiple pieces of the same part or multiple part numbers in this way. Then, to tailor the machining cycle to just the pieces that are actually clamped in the fixture at a given moment, the programmer just clicks on each fixture position to flag that position as either “include in plan” or “exclude from plan.” The result allows the software to produce a multiple-workpiece NC program that reflects the specific workpieces the shop needs to machine.

The same feature also speeds setup on the machining center with the A-axis indexing, tailoring the machining time to the actual number of fixture positions that have been used. The only difference here is that identifying the part numbers in the CAM software will be unnecessary. All of the parts will be programmed in CAM, but the machine itself will select the right set of tool paths after identifying each part through probing its bore.

After The Cut

These considerations related to quickly programming unattended batches and automatically matching the program to the part both relate to the front end of the automated cycle, before any parts are cut. Mr. Gombar Jr. points out that there is another important consideration: the back end of the cycle, which the shop personnel generally arrive to face in the morning. All of the parts machined overnight need to be verified and inspected after the machining is done. In fact, because these are surgical implants, the need to document details of both the process and the part is particularly crucial.

An unattended process that is truly effective for the shop ought to be able to automatically perform this step, he says. In addition to just the time required to perform this step by hand, having human beings making and recording the measurements leaves room for human error.

He sees the machining center probe as key to this step also. Probe measurements of critical features would be sufficient to confirm that the part had been machined correctly—that is, that no tool break, tool wear or fault of the workholding had allowed the part to be machined incorrectly while no one was around to spot the problem.

This information would then need to be stored. The shop is looking at database options for logging and organizing the data as it is reported from the machine’s control via DNC. For example, a seat of product data management (PDM) software could capture this inspection data for each spinal-plate workpiece machined. Such software is more typically used by larger organizations and supply chains, but its database could store all of the manufacturing history information the shop would want to direct to it, including the NC program used to machine that very piece and any e-mails the shop had sent or received related to revisions to that particular part number.

In other words, through further automation of the process, the documentation requirements that are a hallmark of medical-industry machining could become an untouched and even invisible part of the process.

Thus, in a way, the work would have come full circle. The shop won this work by being more than a job shop for one of its customers. Now, by engineering an ever more efficient process around these parts, the effort of producing them will become something much less than what it otherwise might have been.