Rib with constant draft, variable depth.

I'm designing a very large injection molded cover for a piece of
industrial equipment. 58 pound shot, over two feet tall. This cover
needs ribs on the inside top and side walls. We don't want to use
internal action.

Vertical ribs will work, but there is a draft problem. Obviously I
can't just draft the ribs relative to the parting line because they'll
get too thick at the top. So I'll draft them relative to the side
walls, which are inclined between 5 and 10 degrees. The draft on the
side walls vary because they are curved.

This will work because a rib drafted relative to the surface it adjoins
is drafted relative to the parting direction if that surface is
inclined.

What I've done is create a rib that is .200" at its thickest and .080"
at its thinnest and using a variable section sweep referencing the
parting plane, I'm letting the rib depth float so it maintains 1/2
degree of draft relative to draw.

This works great going up the side of the wall. As the wall curves
over, the ribs get deeper to maintain 1/2 deg draft, which is what I
want.

My problem is, I want the ribs on the side and the roof to be the same
rib, but there is nothing preventing the rib depth from approaching
infinity as the rib approaches horizontal on the top.

My question is this: Is there a way to write a relation where the rib
depth (which is being driven by the draft requirement) will limit to 1"
and therefore change from 1/2 degree draft to whatever it wants to be
as the rib rounds the corner from the side walls to the roof?

Does this question make sense?

Thanks,

 

I'm coming up a few pixels shy of a picture. Not your fault.
(Pics and / or model on mcadcentral an option?)

If I understand; I'd sweep the rib section along a single
traj, which represents the rib cap center, with a constant
(perp to pull direction) section normal, trim and solidify.

 

Jeff Howard wrote:

I'm coming up a few pixels shy of a picture. Not your fault.

I'm afraid posting a picture or model is not an option.

That is what I'm doing. However to maximize rib depth on the sides I'm
constraining the rib widths and fixing draft relative to the parting
line to 1/2 degree, leaving the rib depth free to grow as the curved
sides angle in towards the top. The problem is, with no constraints it
becomes too deep on the underneath side of the top of the unit as the
rib rounds the bend from mostly vertical to mostly horizontal. So I
want to drive the rib depth by the draft angle until the rib reaches an
inch deep, then fix the depth to one inch and allow the draft angle to
float. I'm sure this could be done by relation if I only knew how.

Does that make any better sense?

 

I'm afraid posting a picture or model is not an option.

That is what I'm doing. However to maximize rib depth on the sides I'm
constraining the rib widths and fixing draft relative to the parting
line to 1/2 degree, leaving the rib depth free to grow as the curved
sides angle in towards the top. The problem is, with no constraints it
becomes too deep on the underneath side of the top of the unit as the
rib rounds the bend from mostly vertical to mostly horizontal. So I
want to drive the rib depth by the draft angle until the rib reaches an
inch deep, then fix the depth to one inch and allow the draft angle to
float. I'm sure this could be done by relation if I only knew how.

Does that make any better sense?

 

You should consider rotational molding.

http://www.perfmold.com/
http://www.boyce.com/rotationalmolding.htm
http://www.ferryindustries.com/
http://www.blue-reed.com/
http://www.thep2.com/

The size, volume and design challenges fit this type of molding.

David Janes

 

David,

You didn't include my rotational molder www.diamondplastics.com

In general you're right. Large plastic parts that need to be strong are
often good candidates for rotational molding. However, this cover is a
relatively high volume application which would require many rotational
molds to meet demand.

The heat deflection temperature of LLDPE, or even the XPE's commonly
available in rotomolding are too low for this part. PA and CBT are too
expensive etc.

Low pressure gas assist injection molding is a good process for this
application. The process creates a heavy wall and uses an aluminum tool
which is much cheaper than steel. When you consider it would require
numerous rotational molds to achieve adequate volume, the price
difference between tooling options is nearly moot, and the part cost is
reduced by half when injection molded. And the finished part will have
the dimensional repeatability and surface finish of an injection molded
part.

I just need to rib the inside surfaces. I can drive the ribs up the
side as I've described above, referencing the parting plane to maintain
constant draft and allowing the rib depth to float, and use separate
ribs under the top, but the transition will not be pretty between the
two. I'd rather use one VSS and a relation to prevent rib depth from
going over an inch at the top. Do you know how to write that relation?

Thanks

 

http://www.profilesmagazine.com/p21/tips_low.html
This is a presentation from PTC/USER convention that Dave Low gave on
alternatives to the draft feature. Right after that, PTC dramatically
improved the draft command, but not enough to make these techniques
obsolete. He has something in here about how to use a VSS for creating
drafts. Anyway, Dave went from Motorola mold designer to starting his own
company and designed and produced the Skip Doctor, the machine for repairing
the surface of scratched CDs. So, he knows Pro/e and plastic part design.
The article might be of some use.

BTW, do you have a picture of the machine that can shoot a 58 lb part? I
can't even imagine how big it must be. This one shows the Milicron Maxima
with a 607 oz shot (about half of what you need) and 4400 tons of clamping
force. Pretty big, huh! You see the guy on the floor next to it! What's not
pictured here is the robot that comes with it to grab the part so it doesn't
fall on the floor.
http://news.thomasnet.com/images/large/450/450008.jpg Course, maybe you'd do
it with this machine in two shots.

As to relations, yes, with trajpar and a parameter that can be varied, such
as an angle, to maintain the height. It would be some trigonometric
function. I'm wondering though, why not a sketched curve for the top of the
rib so that section will roughly follow the walls but make a smooth turn at
the corner where the wall meets the top. A couple more parallel curves on
the walls/top could govern the .2 thickness at the base of the rib. The
creation of the curve trajectories is perhaps the trickiest part, but I'm
sure it can be done.

David Janes

 

Hmm, I thought I met your guys at a design expo at Chicago's McCormack Place
in the late 90s. Turns out it was Spin-Cast Platics of South Bend, Indiana.

This is good information. I haven't done any design trade studies on the
relative merits of rotational vs injection molding. I'm sure I just reacted
emotionally to a 58 lb part juxtaposed to injection molding when the largest
part I'd ever seen thus molded weighed in at a little over 15 lbs and went
in a 2000 ton press. I didn't mean to advise you on molding methods. I've
been out of the business for a few years and such a thing as a 60 lb.
injection molded part was unheard of (at least, by me) 8 years ago.

Yeah, I know, I just skipped the real question. See next post. Somewhat more
to the point though not necessarily definitive. And, maybe, contrary to your
analysis of what you need, not what you expected. (Though why does the
patient tell the doctor what treatment to give when the patient doesn't even
understand the disease?)

David Janes

 

David Janes wrote:


snip

The machine is pretty large, but probably no bigger than the 4400 ton
press you're talking about.

We currently have a mold at this company where the dimensions of the
mold is around 4' by 6'. That is a big mold.

http://www.horizonplastics.ca/operations.html

This is a page where you can see a 1000 ton structural foam press with
a 260 lbs shot capacity. It is a low pressure process (read: thick
walls). They use an accumulator. 1000 ton clamp doesn't sound like
much, but in a SF type machine it is big, probably an eight foot
platen. These are multi nozzle machines with injection available every
six inches. The part I'm working on is not structural foam, it is low
pressure gas assist, but it is done on the same machines.

This web site is out of date, I'm pretty sure they have a 1500 ton
machine now. In SF, that is very big. It probably has close to a 300
pound shot.

 

The chief thing in all of this is "foam" or "structural foam" and "thick
wall" which means lesser mass (weight), same or greater strength. Hadn't
thought of foam, but, you're right, typical of structural stuff, like the
cover you mention. Thanks for the reeducation.

David Janes