What's Special about SPECS?
That's not the machine I ordered!
By Robert B. Aronson
Sometimes when you've agreed on the price of a machine, the hassle isn't over. How are you going to be sure the machine you think you bought will do what you expect it to after it arrives? For both buyer and seller, misunderstood specifications, or the lack of them, can mean serious problems.
Problems seem to derive from three factors:
- Not enough or too general information. The buyer assumes the machine will do certain things, based on published specs.
- Too much information. The specs are so tight the machine can do a very specific task, but turns out not to be versatile enough to do all the buyer wanted.
- Confusing standards. There are no commonly accepted standards to go by. For example, there are ISO, ANSI, and JIS (Japanese Industrial Standard), and that is often the problem.
Here's a look at how some of these problems can be avoided.
A Government Shop
By Claude Robison
Facilities and Operations Directorate
Oak Ridge National Laboratories
Oak Ridge, TN
When you first set out to buy a machine tool, it pays off in the long run if you put a lot of work in at the front end. Once we decide we need new equipment, there is a general sequence of events.
Determine how much money we have to spend.
Discuss our needs with our contacts at other government labs. We ask their opinions on which are the best machines and who are the more competent vendors.
Use the Internet to see what's available.
Determine what our "drop dead" requirements are.
Determine our "wish list" of what would be nice to have.
Look at the track record of the manufacturers and the vendors that carry the equipment we are interested in.
Review all references.
Begin negotiations. A vendor might say: "We don't have exactly that feature or capability, but we do have something else." And maybe that 'something else' will fit our needs. Or possibly they will cut the price and we can deal.
One important thing--we don't write the specs so tight that we are forced into a sole-source situation. That is, only one company can supply exactly what we want.
Top-of-the-Line Machine Tools
We look at two aspects of performance: noncutting and cutting functions. We do a "spec check" on all our machines, and use the same criteria when evaluating competitors.
By Tim Jones Product Manager and
Karl Lippert, Proposal Engineering Manager
We look at two aspects of performance: noncutting and cutting functions. We do a "spec check" on all our machines, and use the same criteria when evaluating competitors.
Top-of-the-Line Machine ToolsBy Tim Jones Product Manager andKarl Lippert, Proposal Engineering Manager Makino Mason, OH We look at two aspects of performance: noncutting and cutting functions. We do a "spec check" on all our machines, and use the same criteria when evaluating competitors.
For the operator who is looking several years down the road and may need to handle a wider variety of machining challenges, it's often more practical to go to a higher-class machine. That is, one that can be redeployed to handle a variety of jobs. The trick is to select a machine tool that is tough and versatile enough to provide what's needed.
Some companies send in a formal Request for Quote (RFQ) which often has documentation that goes into the finest details of machine tool features and performance. But, in most cases, we don't begin with formal paperwork. A company expresses an interest, and we have to learn what the customer wants to accomplish through customer interviews.
Once we have that information, we supply the customer with documentation that tells exactly what we will provide to meet or exceed the customer's goals. These are the conventional data such as speeds, travel times, power, dimensions, and so forth. We also provide information on when the product will arrive, and all the costs involved.
It's common for a company to create a spreadsheet that is a matrix of all the specifications of a machine and all the machines in which they are interested. It compares the manufacturer-supplied performance data from several machine tools.
But, it's important to know and understand under what conditions these numbers were derived. For example, a figure for "toolchange time" might be calculated in a number of different ways. Do they include door-open-and-close time? Do they give the spindle speed at the time the new tool begins to cut? These are variables that can make a big difference when calculating production cycle times and machine count.
Internally, when we build a new model, we run it through a series of "spec checks" in which we take data directly from the machine's control. Published specs often use some other recording device to develop data.
Although there are formal standards, they are different, and some builders have their own standards. But many supplies still list "done to standards." How were features like acc and dec, maximum rpm, and travel time calculated? Did they use an oscilloscope, or some other device?
Our policy is to take data directly from the control. We test a machine feature ten or more times, then take an average. The customer should only be concerned about what data the controller has. For example, there are cases when a spreadsheet analysis will imply a slower cycle time based on published specs. However, actual cycle times are shorter due to the actual performance and process advantages. It's those data we use when submitting bids.
All our initial tests are related to noncut time. But that's only part of the story. The other half is what happens during the cutting time. We have a separate group of process engineers who look at those issues. This evaluation is linked to specific products, and takes into account product specifications, materials, and production rates. From this information, we tell the customer the best way to use the machine to achieve the results needed.
Our sales philosophy is to focus on problems and solutions, not comparison of published data. We need to know: What are you making? How many? To what tolerances? But we also need to know how the customer is used to working. What type of fixturing and tools are available? What problems have they had in the past? The sale depends on our solving the customer's problems better than the next guy.
Explain the Exceptions
By Paul Vess, Manager
Fuji Machine America Corp.
Vernon Hills, IL
Initially, in looking for the best solution for an application, whether it's a single machine or an entire production line, we analyze and qualify customer needs and specifications. After analyzing a RFQ, we respond with documentation that gives all of the relevant performance information, and covers all other aspects of the machine's design. This includes those related to electrical specifications and government regulations such as OSHA and other safety and ergonomic requirements. It is not uncommon for a builder to take exception to some of the customer's original specification requirements. For example, some requirements may not be compatible or a machine part the buyer wants may not be as reliable as we think it should be.
There are cases in which the customer's original concept of how to best approach a solution may not turn out to be ideal. Fuji engineers are experienced at process engineering, and using the latest technological advancement to provide innovative solutions that might not be apparent at first look. Sometimes this requires an explanation to show the customer the value of a different approach. We then submit a proposal with the optimal solution and reasons why these specifications were derived. For example, we show that a certain performance requirement of a customer might not be cost effective.
Another common problem that surfaces in discussions of customer requirements are irrelevant requirements that are within the customers RFQ. Sometimes these are statements that were written into the document many years ago and never removed, but are no longer relevant.
Once we have agreed on all specifications and costs and we have been selected as the builder of choice, the serious work begins. The next step is design. We generate a detailed set of drawings for the customer's approval. This usually requires some fine tuning, but once all issues are settled, we build. Because of this extensive approval process, lead time is usually four to six months. However once Fuji has the customers specification, build times are considerably reduced; in most cases our deliveries are four months or less.
This extensive process is required for a number of reasons. The two key ones are Fuji guarantees the process and Takt times, and analyzing the fine details is essential to assure success.
A Little Common Sense Goes a Long Way
By Dr. Libor Sedlacek
iFT Institut für Fertigungstechnologie
If you're in the market for a new machining center, there's good news: basically, you can assume that there are no bad machines out there. The bad news, of course, is that a machine can always be applied incorrectly.
Compare purchasing a machining center to, for example, buying a car. If you're in the market for a new car, you can compare models in magazines and using the Internet. Factor in emotion, and you get a result. The machine tool buyer does not have these possibilities. Like a car, a machine tool is a technical product. But there are practically no comparative reports or tests available.
Let's assume we are looking to purchase a high-precision machining center that we will use for moldmaking in hardened steel. Because the range of possible machine tools is so huge, the first step should be to create a list of "killer criteria." This cuts the number of candidate machines to 10 or less. These criteria derive mainly from workpiece size and accuracy requirements. Defining these criteria is simple. First define the largest workpiece that will be milled on the machine tool. Add to this dimension the biggest tool (diameter and length) that will be used, plus travel that the machine needs for acceleration. The result is a total travel requirement for the machine.
If required workpiece accuracy is defined as ±0.01 mm, the positioning accuracy of our machine tool must be less than 20% of this value. Rectangularity can be defined in a similar way. For comparison, it is easier to define this value as a difference that will be measured over a distance of one meter.
Repeatability of such a high-precision machine tool should be five to ten times better than positioning accuracy. Resolution must be the best achievable value that can be reached today. For thermal stability, axes must be cooled or at least held to a certain constant temperature.
And, don't forget that the required precision must be achievable during machining. That means stiffness must be such that cutting forces will not influence precision. Naturally, damping characteristics must also be very high to minimize the impact of vibrations on workpiece surface finish and tool life. Hydrostatic guides should be preferred in this case.
This should result in a substantial decrease in candidate machines. If more than 10 machine tools remain in contention, tighten the accuracy and other requirements.
Valid comparisons require correct data from machine builders. Builders that cannot or will not provide the needed information are eliminated from the competition.
The goal is to determine three finalists, which will be subjected to metalcutting tests. We need a standard test that must be well-known by the machine tool builders.
For this, we recommend the application of ballbar test equipment. You need results of interpolation of circles with the same scale and radius. Often a visual comparison of circles one below the other will allow selection of the "best" three machines; otherwise, analysis software used with the ballbar will aid in selection.
Next are machining tests that evaluate the dynamic behavior of our future machining center. Allow a minimum of one day for each machine tool.
Be sure to take two experienced shop employees, who can provide their impressions of how the machine is functioning during cutting. Also, take enough work material for testing. The material should be certified, and ideally should come from the same heat. Machining data should be prepared in advance using your CAD/CAM system, and all tests should use the same workholding method.
For tooling, be sure you have a sufficient quantity of the required tools, and they must be the same size, same grade, etc. Equipment needed for evaluating test results may include surface roughness testers, a digital camera, a microscope, and a stop watch.
The actual test methodology will depend on you. A nice example is milling of hardened steel up to RC 65 with surface finish of Ra = 0.2 µm. In this case, selecting a machine tool may be very difficult--only a few builders have experience at this level of technology.
Cutting tests can vary. Common ones include machining a helix or circle-diamond-square. Our preferred method is milling of a planar helix with a pitch of 0.1 mm and machining tolerance of 1 µm.
The cutting tests will generate a lot of data. However, three criteria are of prime importance:
- Speed--the shortest time needed to machine one helix length.
- Stiffness/damping--the best workpiece surface finish.
- Stiffness/damping--the longest tool life.
If no machine can fulfill our requirements, our expectations regarding price and performance were unrealistic. But, if all three machines meet our criteria the machine that needs the shortest time to machine a given helix length will result in the best profitability. Many people think a machining center moves according to the programmed feed rate, and this is true when movement is in a straight line. In curves, however, the programmed feed rate will not be reached. In our experience, the difference between machine tools and controls in this regard can be quite large.
For this reason, the measurement of reached helix length is very important. Particularly in hard milling, there can be big differences between machine tools, and the length of the helix is a function of the machine's stiffness and damping characteristics. We were surprised to learn that machine structures without good damping can produce helix lengths orders of magnitude smaller than stiffer machines.
Evaluation of cutting data provides a technological and economical basis for selecting a machine. Other factors, such as quality of service/support and delivery time, are still not included.
These factors can be used for the final decision, or can be taken to the pre-evaluation depending on their importance to your company.
Buying Five Axes
By Gary Zurek
Applications Engineering Manager
Mikron Bostomatic Corp.
When we have a customer who is interested in our five-axis machines, we commonly will take the first step by completing a time study of their part. This includes process analysis, fixturing, selection of cutters and tooling, programming strategies, etc. It includes everything except the physical cutting process at the machine.
This does two things. We get a good idea of the performance achievements that we can bring to the customer, and the customer has a good idea of what our machines can do along with the costs, cycle times, and tooling investments involved.
Once we agree on the time study, the final test is actually cutting the part to the customer's satisfaction.
Test Parts Tell the Tale
By Leonard LaVoy
Farmington Hills, MI
(Vicount Industries is a 60-person, machine-tool company specializing in progressive and transfer dies, chiefly for automotive customers.)
In general, we've had no major problems with our machine tool suppliers, aside from minor software glitches. But in our selection process we first, through experience, know most of our key parameters--such as horsepower, spindle speeds, and the like. Then we do a lot of research, and usually narrow the search down to two or three machines. Only top-of-the-line machines are considered, such as Toshiba, OKK, and Okuma. We don't even look at the commodity machines.
The final test is to make some sample parts. When we go for demos we rarely pay attention to what they show us. We want them to make our parts to our specifications.
We give our potential supplier our test parts early on the test day, and ask for the finished part the same afternoon or the next day. That way there is less chance for the supplier to do any "special tweaking" of the machine tool or programs, and our samples come out close to the way we would end up making them.
After we narrow the choice down to a couple of machines, it becomes a bidding war. But performance can be more important than price. If the difference is only $15,000 or $20,000 you are quickly going to eat that up if the machine doesn't perform. When we buy it's usually for expansion, but most recently our purchases have been to replace older manual equipment with three-axis CNC units.
All the major manufacturers produce quality machines, so you have to determine which machine best fits your needs and performance requirements.
There is a constant seesaw battle between the machine-tool builders and toolmakers. First one is ahead, and then the other. Right now, I'd say it's the toolmakers' turn to catch up.
Buying EDM Equipment Is Different
By Brian Pagano
A company may stay with a single supplier more out of tradition or familiarity with the product and the company. As a consequence they may not stay aware of advances in the industry. This is particularly true of EDM. Unlike traditionally slow moving conventional machining practices, there have been major leaps in the technology since it was first introduced. For example, cutting speeds have gone from 2.0 in.2 (13 cm2) per hour in 1" (25-mm) thick steel in the late '70s to today's speeds of 46 in.2 (297 cm2) per hr in 2" (51-mm) thick steel.
There have also been major developments in finish and corrosion elimination to be aware of. Initially, EDM chiefly served the tool, die, and mold industries. Today, finish and cut complexity are a strong competitor to some conventional machining operations.
What About Training?
One issue with the more complex multifunction machines is: Who will run them? Reports from users and manufacturers vary. On one hand some suppliers worry that customers, particularly first-time buyers, don't have the training to take advantage of all the machine can do. Then they end up blaming the seller because they are not getting the results they expected.
This can sometimes lead to an extensive, costly, hand-holding time for the seller, or the seller has to be capable of supplying training as part of the sale. On the other extreme, some sellers find that first and second-tier manufacturers often have more competence than operating personnel in very large companies, such as auto manufacturers. Tier suppliers may be more motivated and have a wider range of skills because of the variety of jobs they have. Some machine tool builders have found that major manufacturers sometimes want to factor the operator out of the work cycle as much as possible both because of high labor costs and the chance of introducing error.