- THE MAGAZINE
- WEB EXCLUSIVES
I’m not much of a gambler and fortunately for me, there was little time available to make an even larger contribution. But one of the frequently asked questions at the meetings got me to thinking about the gambles many manufacturers make every day.
During the meetings we were reviewing a number of standards and, in particular, specifications regarding thread gages. There were many different requirements listed along with their tolerances and, more than I’ve noticed before, participants were asking, “How are you going to measure that?”
Far too often manufacturers spend a lot of time designing components without this critical question being asked. And in many cases, there are few on staff knowledgeable enough to make suggestions that might work or to point out those that won’t. Those companies that do have such people tend to get them to review specifications and drawings before the chips start to fly. Those that don’t will get the message when the parts won’t assemble or a customer rejects them.
Because so many factors must be considered when selecting equipment, looking through supplier catalogs will not be too helpful. For example, let’s say you’re dealing with a simple ground shaft that is 4 inches long and you have to verify the outside diameter. Fixed limit ring or snap gages could do the job but if there is a taper limitation they wouldn’t be of much help. A standard outside micrometer or dial snap gage would handle this situation assuming the machine operator was skilled enough-and had enough time-to use them.
On the other hand, if there is a straightness requirement, other equipment would be required. If there was no significant taper requirement and the outside diameter tolerance was suitable, go/no go plain ring gages 4 inches long would ensure straightness stayed within the size tolerance. But if the straightness requirement was closer, the rings won’t help no matter how long they are.
If the tolerance on the diameter is tight, you can pretty much ignore the previous paragraph and go back to square one. If this is not a problem but roundness could be, the best micrometer in the world won’t reveal what needs to be known. Having beaten temperature variations to death in a couple of recent columns, I won’t bring them up here, but they also have to be considered.
Some folks take the easy way to deal with such situations-one that doesn’t involve all these messy technical details. They simply find out what their customer is using to measure the parts and then copy it. The fact that it may not do the job correctly becomes irrelevant because, hopefully, both parties will get the same “readings.” Of course, if they get different readings someone will have to find out why.
Another way to skip the technical stuff, particularly if your customer has equipment you won’t spend the money on, is to have the customer measure a few parts. Comparing those measurements to yours, you may spot a trend. You could make parts to the wrong tolerances by your measurements so your customer will find them to be satisfactory.
Engineering often puts tolerances on drawings that are tighter than required to ensure they get what they want, which adds another dimension to the game-no pun intended. Accurate measurements often reveal that these pretend tolerances were never met anyway but the product worked as required. This leaves a legacy of belief that mystical tolerances are being maintained that will carry on until work is undertaken for a customer that can measure the parts accurately.
Unfortunately, these games cannot be played for long before someone gets wise and brings the down the whole house of cards. And then someone will have to look into those messy technical details for the answers and let the cards fall where they may.
I prefer not to gamble like this because it’s so uncertain and risky. At least the five cent slot machine is very certain: It wins over time no matter what strategy you think you can beat it with and in my case, I’ll fall asleep before significant money is at risk.