Case Study: Managing Conflicting Constraints

I’m looking for a good source of Unobtainium.  With the pandemic in full swing, it’s really hard to find.  Yet, the demand keeps growing.  I’ve got a few customer projects that need it to solve conflicting constraints, so if you know a source, please pass it along.

I think Unobtainium would really have come in handy on a recent project with conflicting engineering constraints.  You’ve likely been there too.

Today’s example is a simple structure.  The customer originally approached me with a few quick questions about cutting steel.  As we got into it, I realized is was a lot more.  While the project definitely needs cutting, it really needed a little more thinking about how to accomplish the broader picture goals.  Since the big picture really sets the stage, what started out as simple questions turned into a project.

The Big Picture First

Remember the Triangle of Achievable Engineering?  Yeah, that one where we talk about hitting objectives.  Here’s the triangle.

So, it is a useful tool when we think about other goals too.  For this project, we can relabel the sides with the goals of Weight (lighter is better) / Vibration (minimize it) / Accuracy (for positional interaction).

For this project, the customer has two small machines that mount on the two pedestals.  Both machines create some vibration, yet the two machines must interact solidly with respect to each other.  Now the hiccup, the assembly must be light enough to hand carry.

Looking at the triangle . . .

• It’s easy to get Low Weight and High Accuracy if we drop the Vibration requirement.  Accuracy can be great statically, but enter the vibration on a lightweight platform and accuracy goes out the window.
• It’s also pretty easy to make it rock solid even with vibration if we don’t care about weight.  A huge solid chunk of machined steel would certainly do the job, but much to heavy!
• The combination of low Weight with well damped Vibration is also pretty easy if we don’t care about accuracy.  Rubber mounting does that, but it would move around and fail to achieve any kind of accuracy.

You get the picture.  It’s all about managing conflicting constraints.

A friend once described engineering as The Science of Balance.  It’s true in physical ways, human ways, mental ways, financial ways, and so many more.  No wonder we reach so often for the holy grail of something we can’t have.  ex. Unobtainium.

Bringing Things Into Focus

Well, if I can’t have the infinitely stiff, infinitely strong, Unobtainium that weights virtually nothing, then I’ll have to settle with another method.  Innovation is my “Go-To”, what is yours?

For me, the Elegant Solution usually comes as I shift paradigms.  Some people call it thinking outside the box, but I don’t really recognize a “box” per se.

This customer came asking about cutting steel (angle, tube & flat) to get the alignment right.  My first reaction was one of accommodation — until I really understood what he is doing.  As the higher level goals came into focus, it became clear the solution was not down the traditional path.  Yes, it’s a big-ish pedestal, and welding a structure is good, but a different approach will come closer to reaching the big picture goals.

The Solution

Let’s ditch the structural steel, and instead, make a 3D puzzle.  A bunch of interlocking plates will make assembly easy.  It will also achieve a fairly close alignment — almost by default.

However, the 3D puzzle pieces can’t be the alignment.  For good alignment, we need fixed datums that are datums by definition.  Puzzle pieces can interlock, and provide the end strength, but the cut edges do not make good datums.  That will end up a mess.

Welding will join all the pieces and solidify the assembly once the assembly is complete.

Some internal pieces (the cool tid-bits, not shown) create the datums and all but assure a square, aligned fit.  (That’s really the magic for success with this project, and it reflects a primary reason to hire Synthesis.)

The broad plates will carry all the forces crisscrossing within, and give great surface area for the internal damping material — sand.  (Loose sand is an awesome damper for vibration.)  Finally, if the damping material (sand) is put in and taken out separate, then it becomes modular (in a different way).

That gives a pedestal with a reasonable balance of the goals.  It is both light to carry, and heavy (plus) for damping any vibration.  It also achieves a very stiff and accurate pedestal for the machines.

Truth be told, because of the inconvenience in handling sand, he will try it without the sand first.  Then, use the sand if we find the vibration objectionable.

Did We Balance The Conflicting Constraints?

Looking at the triangle — Weight, Vibration, Accuracy — how did we do?  We didn’t compromise much on Accuracy, and we found a good strategy for weight and vibration (function), by increasing cost and inconvenience.  (Handling sand is definitely an inconvenience.)

So, really our paradigm shift changed the triangle.  We end up with triangle legs of Accuracy / Weight & Vibration Balance / Convenience & Cost.

Of course, there are a lot of ways to think about the triangle, and maybe you see it different.  The point here is a change in what we first think can lead to solutions that are quite different from the original vision.

When Is Innovation Innovative?

Is that a silly question?

There is nothing very innovative about the solution.  It’s just applying differently some things we’ve done elsewhere.  A new problem, with several old solutions mashed together for something that looks kind-of new.  Is that really innovation?  Well, I have a lot of patents that are basically just that.

So, for this project, you can see the images.  By applying principles of care from the design, to the parts, to the build, we came very close to parallel and perpendicular just in the weldment.

Starting in the engineering, we applied good sense in what we could expect for part tolerance in ways that gave dependable datums for building.  We also thought carefully though the manufacturing processes and biased the design for methods we knew are easy to make right.  It’s one thing to envision something perfect, and quite another to actually put the parts together.  Again, a different kind of conflicting constraints, because you have to build it.  That included leaving appropriate access — in this case to weld on the inside.

To answer the question, it can feel innovative when it all comes together and delights the customer in ways that they were not expecting.  Which is not so different from the discussion on Prototyping the Impossible.

So, the twin pedestal, 3D puzzle box is light enough to lift, and it’s amazingly stiff.  Plus, it has all the needed adjustment so accuracy is achievable.  In truth, we did sacrifice some on weight and on convenience to move, but we hit dead on with accuracy and vibration.

The Elegant Solution from Conflicting Constraints

When it comes right down to it, there is not a fixed path for achieving engineering balance.  Getting out of a thought groove definitely helps.  Changing the paradigms of conflicting constraints (whatever they are) also helps.  Seriously, creative thinking is often better than Unobtainium.

But, I think the most useful piece is a willingness to go back to the big picture and look at the world different.  Never settle for the first possible solution, and stay out of the weeds until you know the furrow.  (Sorry, I’m just an old farm boy.)

Good Luck With Your Projects!

Share the Article: