Jump to content
Mechanical Engineering
  • entries
  • comments
  • views

Is Form Fit and Function Sufficient for Interchangeability?

JAG Engineering LLC


My answer to that question would have been yes, and like all absolute answers, it would prove to be incorrect. An axiom in manufacturing is that the design engineer should not dictate manufacturing processes unless it is vital to the function of the design, metal vs. plastic, casting vs. forging etc. Once passed the gross requirements, dictating which machine tool to use would be beyond the design engineer’s prerogative unless there was a compelling reason! However, this should not be taken as the design engineer not learning all he can about manufacturing.


Requirements come in three major categories. Those we know need special attention, those we expect to happen as a matter of course, and those we did not know existed. If we need a very tight tolerance, we will call it out in the documentation for a specific feature.  For the second category, we rely on title block tolerances or good machine shop practice. Those that fall into the third category are discovered when a product has been in production for a while and then a seemingly benign change is made. Often the original design team is no longer in place. Whether the original team thought a requirement in the third category was covered by the in the second category may not be possible to determine.


Products go through a test phase to weed out the unknowns before release, but revisions don’t always get the same scrutiny. Here is an example where after hundreds of thousands of units produced a change in manufacturing process caused a failure.


The product was a servo assembly that has a 200:1 reduction. It was smaller than a pack of cigarettes and the small DC motor had very little torque at the motor output shaft. I think a person could stop the shaft by hand. Two hundred reductions later there was no way you could prevent rotation at the output of the servo assembly with your bare fingers.


Some changes in the manufacturing facility resulted in a turned part being produced on a mill. The part can best be described as an axle with a flange. When the part was turned the figure on the left resulted. The concentric lines represent the fine spiral grooves on the face of the flange produced by the cutting tool when the part was turned. The figure on the right shows the fine swirl marks from an end cutter that circled the axle projecting out of the view towards the reader.


The surface finish on the flange face was never specified having it fall into category two. How much friction resulted was never established or never documented because turning produced an assembly that worked. The consultants who designed the servo at the start up of the company were long gone. The new surface condition produced by the end mill increased the friction to the point that the servo would stall. The surface in question was close to the motor so the torque at this interface was not much more than the motor output and a long way off from the 200:1 final reduction.


There are a few lessons here:1) minor changes are not always minor, 2) good shop practices are a risky thing to rely on,  3) when something stops working look for the third category of requirements, 4) ask yourself what exists in a part that is not defined or inspected, 5) the design engineer needs to be close to manufacturing even when he needs to remain silent on process selection, 6) manufacturing should notify engineering when changes are made to a process that has been in place for a long time.


I wish I could claim credit for discovering the cause and fix but a very sharp manufacturing engineer named Bob C residing at headquarters identified the cause.



Recommended Comments

This is an excellent example of real-life engineer problem solving. We could look at the two parts, one manufactured on a lathe and the other on a mill, and say that they appear to be the same. Only rarely do we get down to the level of inspecting the surface finish marks, and yet, in this case, it made a critical difference.

Thanks for a great post.

Link to comment

This, in my opinion, reveals the need for written, general manufacturing standards to be adopted by any design function which may not have direct oversight in the manufacturing process. these standards could then be submitted to vendors as a go to standard in the absence of specific guidance.  It seems like a heavy burden to begin to create such a document, but including titles such as: "Minimum Surface Finishes for Dynamic Interfaces" could lead to someone at least having to ask the question before changing the manufacturing method.

Link to comment

A very pertinent example by JAG that highlights one of the results of process-change in engineering production when communication between engineers is absent. Surface finishes formed by different machine tools are quite apparent in cross- section when viewed through a microscope, but not necessarily to the naked eye. Components that are produced on different machines, such as the lathe or milling machine, whilst they may adhere strictly to manufacturing tolerances, can offer totally different characteristics in operation such as variations in the frictional resistance referred to by JAG.

Production engineering sometimes makes advancement by the 'suck it and see' method, but hopefully engineers will take heed of this post and always consider the surface finish that different machine tools leave behind.

Link to comment

This highlights a point I make that there are no "routine" or minor changes until you have done an off line test of the change you wish to implement.  There will always be times when you don't know that you don't know, so do your learning while sitting at the kid's table and only implement changes after due diligence is done.

Link to comment

Very true. Often we are luckier than we would like to believe. I have written one article that touched on incremental changes that slowly pushed us to the envelope of a design parameter. Each change further lulled us closer to a cliff. In Russian roulette each try you are lucky gets you closer to being unlucky.  

Link to comment
Add a comment...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.

  • Join Mechanical Engineering network

    Join us (login) to get full access : Please sign up to connect and participate.

    To download files...please login

  • Create New...