Thanks for this great resource!

Harald

If you want a better understanding of how plastic parts are designed then head over to efunda and have a read of the “Plastic Design” category under the “Design Center”. They do a very good job of detailing plastic part design guidelines. There is quite a lot that goes into it.

Material thickness is determined by the type of plastic and how I intend for the part to function. Part function is most important but proper wall thickness and coring have to be taken in account so that the part will mold accordingly. Applications like MoldFlow help me to analyze how a part will mold before I have a tool built. I watch for how material flows, where knit lines occur, how gate locations affect things, and how the part will shrink. Also, I can check for proper vent locations so that gases during molding can escape without interfering with the integrity of the part. Improper venting causes voids inside of the part rendering it weaker in that particular section. MoldFlow can help a designer (me) focus on the part while a mold maker can use it for proper tool design. Also, I can perform some light finite element analysis (rarely do I, anymore) if I think it’s needed to check part stresses (flanges). FEA is a science all to itself (I’m sure you know) so if you do not understand how to apply proper loads then you’ll get false information. It’s not my area of expertise so I try to use other methods to prove out a concept.

Proper flange design can be a little tricky. Sometimes it takes a little math, a little analysis in CAD, a little prototyping, and a little experience (meaning failure so that I know what not to do). Making mistakes can be costly after a mold has been made. Just like designing buildings, I suppose. Takes a little foresight to minimize those easter eggs.

Prototyping always starts out for me with a virtual concept in 3D in my CAD system. Once I have settled on a particular design I may choose to make a prototype. Part complexity determines prototype method. For instance, I may go with an SLA model from a 3D printer just so that I can evaluate size and function. From there I might make some modifications and prototype again or go straight to final tooling which will be a multi-cavity injection mold. Another method of prototyping might be a single cavity MUDD mold which is small and relatively inexpensive to build. I like to have things prototyped but I also have to be practical about cost. I also make my own prototypes at work. I have a mill and a lathe and I use them often.

As far as keeping the part from leaking, much goes into that as well. Proper o-ring/gasket design is involved. Take a look at efunda again. This time o-ring design guidelines.

Before I begin concepting a product I tally some guidelines. What must the part do? Who is it for? How much should this part cost? What color should it be? What material is most suitable? How will this part be made? Who’s turn is it to bring in donuts tomorrow morning? Is there a fresh pot of coffee? Did I wear these underwear yesterday? Is that a booger on my finger? What stinks? Is Verbal going to send in a listener tip this week?

How do you decide the thicknesses of the saddle (like the flanges and main body of the saddle). Do you just pick some reasonable thickness, fab a prototype and try it out? I mean if you picked something too thin, it might be easier to install but wouldn’t hold as well to the extrusion – it might leak. How many prototypes do you try out? Or maybe you’ve got it down to a science. (ie. you do it with mathematical models etc.)

Shed some light for a know-nuthin’ like me.

Rendering

By the way, what kind of tool is it anyway? I just cannot figure it out.

ps. Your fingers don’t look nearly as sausage-like as I had imagined. Did you get a manicure done specifically for this?