Using stereolithography for low rate manufacturing of plastic covers 3

[KENAT]

Not sure this is the best forum but here goes.

We manufacture precision metrology tools, with run rates in the order of 50 or so per year, some slightly more, some a lot less. Historically the life cycles have been fairly long, 10 years or more but that seems to be reducing significantly.

On our latest designs we’ve been making use of plastic (the drawing I looked up said polyurethane with no grade specified sorry this is a bit vague) to improve the aesthetic appeal of our product. While the individual covers vary in size think of things from the size of a food processor or sewing machine (cause that’s what the first product of ours that used plastic covers looks a bit like to me) up to a large cash register or maybe 21” CRT size.

To me a major issue on this is the cost and timescale involved with making the molds for these parts. Our suppliers typically make a temporary silicon mold for the first 10-20 to prove out the design etc and then we go to an Aluminium mold. The silicon molds are typically several thousand $ while the aluminum molds are several tens of thousands. The molds take weeks or even months to get made too which is a problem as we’re trying to reduce our development cycle.

In order to reduce cost (at least up front cost) and speed development I wonder if these covers would be a good candidate for some kind of ‘rapid manufacturing process’ like STEREOLITHOGRAPHY or SELECTIVE LASER SINTERING.

I had a quick search on the web and this site (thread404-10501 for instance) but had trouble separating the relevant information from the various patent claims and minimally informative articles etc.

Any advice on the current ‘state of the art’, relative costs, potential sources or information or even vendors etc. would be appreciated. I’ve dealt with stereo lithography parts before but they were far to fragile even for our relatively delicate instruments.

Thanks,


KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

What is Engineering anyway: faq1088-1484

 

[ornerynorsk]

We have a Dimension BST machine, about 5 yrs old now. It uses the FDM process, ABS material. Their newer machines lay down deposits with a material/fusing strength of 40% stronger than the older machines. We make parts all the time for prototypes which go out for trial eval, trade shows, and the like. The ABS material is quite strong, and compatible with most paints and fillers you will find in an automotive store. The parts can really be made to look nice, as well as being structurally acceptable. We find that the cost of doing it inhouse is about 20% of having it hired done. Just my 2 cents.

 

[Demon3]

Stratasys also makes machines that will print parts from molten thermoplastics such as ABS, polycarbonate and polyphenylsulfone. The cheapest machine is around 30 000 Euro I believe. The parts will have decent mechanicals but poor surface finish.

Chris DeArmitt

"Knowledge has no value except that which can be gained from its application toward some worthwhile end."
Think and Grow Rich - Napoleon Hill

 

[patprimmer]

I think casting resins in cheaper cast aluminium or medium density fibre board moulds would be your best choice at those volumes.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[Demon3]

I have seen such machines used to print an injection mold which you can then use to make thermoplastic parts in limited runs. Amazing and cheap compared to metal molds.

Chris DeArmitt

"Knowledge has no value except that which can be gained from its application toward some worthwhile end."
Think and Grow Rich - Napoleon Hill

 

[ornerynorsk]

A Dimension machine starts at about $22K USD, if memory serves, FYI. That's with a 8x8x12 inch build envelope. They have larger machines as well.

 

[ornerynorsk]

One critical detail, which I nearly forgot about, but is worth passing along >>> It is quite easy to section larger structures so that one would build multiple "pieces of the puzzle" if you will, and solvent cement them together into one structure. We've had much success using our small machine to build structures as large as 22 x 24 x 4 inches in this fashion, and there really is no limit. Additional structural ribs, which can be solvent cemented into place upon assembly, are easy to add, as well. This keeps the capital expenditure down, going with a smaller machine.

 

[Demon3]

For your information, although the standard materials are ABS ad PC, I did some work with Stratasys a few years ago and we made nice parts with MABS (transparent ABS) and ASA (like a super-weatherable ABS).

Chris DeArmitt

"Knowledge has no value except that which can be gained from its application toward some worthwhile end."
Think and Grow Rich - Napoleon Hill

 

[pierdesign]

Hello.

It's a good question, often repeated at companies such as you describe.

Unfortunately, rapid prototyping processes parts still are not acceptable for long term performance on high value added instrumentation.

All these various processes, while very promising and subject to what I think is some spurious marketing claims, have issues that compromise long term performance: usually evidenced by breakage at assembly interface points - stress.

Essentially, all these processes, properly still carry prototype in their name. This is where the spurious marketing comes in, they've dropped it in favour of manufacturing; it's misleading. Of course, there are some success stories of production parts, but I believe these are non demanding performance assembly or stand alone piece part applications.

For companies as you describe, I suggest two methods to address issue. Pressure vacuum forming or reaction injection molding. With some intelligent design, molded parts can be leveraged in future product iterations with specific areas further machined and accesories asembled in place. RIM is ideal for this as the process disregards constant wall section, any wall thickness is possible. Hence, in design, specific areas could be targeted as customizable, and additional details, wall sections, etc, designed into the tool for molded parts to be machined later. RIM tooling is about 10 - 30% of injection mold tooling, for reference.

 

[patprimmer]

I agree with pierdesign. I forgot about RIM. It is a good process ad is often overlooked. Parts with good properties can be made. Thermoforming of sheet by various methods from vacuum to draping over a mould and pressing down with leather glove covered hands.

Two pot casting resins such as Epoxy, Acrylic, Polyurethane or Polyester can also work.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[ornerynorsk]

Kudos to pierdesign, yes we get a little short sighted sometimes!

 

[HDS]

The parts made by processes that have evolved out of rapid prototyping into rapid manufacturing can work for some people. However, the part always has to be designed for the process. The trade off of rapid manufacturing is that part prices are higher but there are no tooling costs or lead times. Stratasys is trying to become the leader in this.

Your description sounds like you are already using RIM. As everyone else has said look into thermoforming. It can be very low cost and low volume but may have less detail or features. Look at thermofab.com and thermopro.com or

http://www.thermoformingdivision.com

for examples.

Fiberglass can be an option for the larger parts.

 

[KENAT]

Thanks for all the responses, it looks like I may be premature in considering derivatives of rapid prototyping for actual manufacturing, although one or two voices are more optimistic.

I believe our current vendor is using the RIM method though I’m not totally sure. We have looked at vacuum forming but most of our parts are too complex.

One thing I forgot to add in my OP is that one of the other problems we have with our molded parts is that none of us has much experience designing molded parts. Some of our designers seem to have real trouble with some of the design for manufacture concepts like including draft etc. and I’m not that much better but have got involved where they were having trouble. The design process for the covers on our last product was painfully long, months in fact. I was hoping using stereo lithography or the like would reduce these problems as well as tooling cost & lead time. We tried outsourcing the design of the covers but this was even worse.

I'd found one potential lead for rapid manufacturing

http://www.redeyeondemand.com

but based on most of my other reading I’m not overly optimistic as many of you seem to agree.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

What is Engineering anyway: faq1088-1484

 

[btrueblood]

Kenat,

Any chance you could post an example picture/sketch/drawing?

 

[KENAT]

I held off on posting images but here goes, hopefully no confidentiality rules breached.

The left two are new parts on a new product, I’m not sure what the plans for them are at the moment but they were what prompted me to look at this again in more detail. (Looked at it briefly a while back but came to the belief the technology wasn’t mature enough.) On the top is about 4” long, on the bottom about 6” long. These are pretty simple compared to some parts.

The top right is a ‘cover’ that has several parts mounted to it. There are loads on it as some of the parts mounted to it are PCB’s with connectors that get pushed on/pulled off. Most of the holes have threaded inserts in them. It’s about 15” wide (longest dim). The bottom right is about the largest plastic part I think we have, something like 24” square.


KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

What is Engineering anyway: faq1088-1484

 

[patprimmer]

Kenat

Have you approached the major resin manufacturers like Bayer, Ticona, DuPont BASF for their design guide litrature.
If you need some good practical designers, I know a few. I expect there are a few who post here as well.

Regards
Pat
See FAQ731-376 for tips on use of eng-tips by professional engineers &

http://eng-tips.com/market.cfm

for site rules

 

[btrueblood]

Kenat,

It looks to me like the two left-hand parts are good candidates for thermoforming, possibly with some tweaks to add draft, and post-forming machining/trimming to get some of the details added.

The right hand two parts...it looks like there are some areas with very thick material, unless we are not seeing the backside is indented. Fat sections are generally a no-no for plastics, except for RIM which (can?) create a foam core...but that causes trouble for threaded inserts which I also see there. 'Twere me, I'd redesign the two right side parts to be multi-part assemblies.

I don't have a lot of experience with plastic injection molding design, but a fair bit for metal castings, powder metal parts, forgings, extrusions, and rubber transfer and compression molds. Oh, and thermoformed plastic parts. Oh, and some experience re-designing plastic injection mold tooling to fix troubles. ALL of the above experience was gained most rapidly by visiting the shop that was doing/going to do the work, and finding out from them first-hand what the limitations of the process were, and where the draft would have to go, etc. I highly recommend such visits for designers, esp. if they are having trouble designing parts that work right the first time. Knowing what your vendors CAN do, helps in designing parts to suit their expertise and methods.

 

[KENAT]

We actually got some reasonable design guidelines from our vendor, but other designers just didn't seem able to get their heads round some of it. Plus, they didn't build the models from the start to really allow for moulding, more to suit a woode mockup one of the guys built in his garage, and then tried modifying them.

Some of the parts, definitely bottom right, were originally multi part designs but our vendor suggested making them one piece to save cost.

Thanks for the feedback.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

What is Engineering anyway: faq1088-1484

 

[pierdesign]

it's easy to criticize....

There are some plastics design issues with your parts. Sharp edges, large open areas, unsupported tabs. But, I'm sure they're still in development.

Regarding RIM, Premold in Wisconsin has an excellent design guide

http://premoldcorp.com/designguide.php

(no, I don't work there)

It can be difficult to comprehend some manufacturing intracacies. Certainly a factory visit to observe both tool construction and molding process can be very helpful.

 

[KENAT]

Just realized I didn't explicitly explain the two parts on the right are already in production and were 'fine tuned' with the input of the vendor. In fact, I just re-read part of your post pierdesign and I think they may actually be our current vendor.

KENAT,

Have you reminded yourself of faq731-376 recently, or taken a look at posting policies:

http://eng-tips.com/market.cfm?

What is Engineering anyway: faq1088-1484