Friday, May 27, 2011
Monday, May 2, 2011
Wednesday, March 23, 2011
Thursday, January 20, 2011
Update: Carbon Fiber Compression Molding of Centrifuge Cups
For all the masses out there at the edge of their seats waiting for the next in the series...Our last post documented our process of ramping up from prototype to production of our carbon fiber centrifuge parts. A quick recap on the project scope- our company, Apex Designs, was hired to develop and manufacture lightweight but exceptionally strong cups to house viles or bags of blood that are spun in centrifuges for medical research. We proposed to build the parts from carbon fiber- specifically bi directional pre-preg carbon. Because our client required a smooth finish on both the inside and outside of part, we opted for a compression molding process that would give us tooled finishes on all surfaces of the part. (See previous blog below on the development of the tooling.)
We use a CAD software called Solidworks for our design and simulation work. Solidworks does a very good job at performing stress analysis on parts and assemblies made from most of the
common materials used in manufacturing- metals, plastics, etc. It does NOT however, allow you to perform stress simulations on composite materials (well, not without an add on $12k non linear simulation program). So if you are designing parts using composites, you're SOL on getting any help from the computer. However, it is still very helpful to perform an analysis on the part by assigning other materials to the part just to be able to see the highest stressed areas or features of the parts even if the numbers dont make any sense.So, in the real world, good old hand calculations are used to give you an idea of where you think you should be and testing is required to verify your calculations. The cool part about composites is that you can orient the fibers of the layup to give you the strength characteristics that you want. The reality is, however, that testing is absolutely necessary to validate what you think is the proper layup schedule. Because we are building parts that have an aluminum ring inserted into the layup- and did not know how that would effect our part's strength, we sent a few samples out for testing to a proper engineering materials testing lab. The tests results showed our prototype cups met the strength requirements set forth by our customer, but because we were going to build several hundred parts, we decided to build our own testing rig to test samples of the batches built in house. This would allow us to verify the consistency of our parts and allow us to sleep at night!
Straight out of college, I worked as a materials
test engineer where we conducted all kinds of materials and product testing- we got to break to alot of stuff- everything from metal roof panels to 25 ft long fiberglass columns under hundreds of tons of load. My background there in ASTM standardized testing gave me the confidence that we could build our own test rig.
To load up our parts, all we'd need is a simple hydraulic ram with a known piston diameter, a good quality pressure gage, and a nice rigid frame to place the cups under load. We converted a 20 ton hydraulic press into our test rig by adding a rigid fixture to hold the cups, and adding a pressure gage to the ram. For peace of mind, we dis-assembled the ram to measure the actual piston diameter instead of just assuming the literature it came with was correct. It was as stated, but if it were wrong, our calculations would be totally bogus.
Pressure = force/ area where pressure is in PSI, force is in Pounds, and area is in inches^2. Figuring the force applied to the cup is a simple matter of converting the pressure registered on the gage and multiplying it by the area of the ram's piston. For our tests, we loaded the cup in 200 psi increments and held the load for 30 seconds before proceeding up the scale. In most tests, a strain gage of sorts is used to measure the deflection of part at each load. The simluation from our Solidworks software showed us the most likely type of deformation, so we chose to measure that deformation using dial calipers. The good news is that the deflections we saw were pretty miniscule for the loads applied. Carbon's yield point and ultimate failure point are very close together so the lack of deflection was really of no surprise to us.
With the testing rig in house, we are now able to validate our part's strength characteristics and it will allow us to try new methods or layup schedules with new products that are surely to follow our medical centrifuge cups. And no - you dont get the numbers and no I'm not ready to divulge how we solved our silicone tooling problems- well not yet anyway! Like my friend Scott says, "a baker's success is in how he protects his recipe" Of course, if I can be of any help, I'd be glad to speak with you. Our contact info is on our website http://www.apexdesigns.net/
-Steve
Straight out of college, I worked as a materials
test engineer where we conducted all kinds of materials and product testing- we got to break to alot of stuff- everything from metal roof panels to 25 ft long fiberglass columns under hundreds of tons of load. My background there in ASTM standardized testing gave me the confidence that we could build our own test rig.To load up our parts, all we'd need is a simple hydraulic ram with a known piston diameter, a good quality pressure gage, and a nice rigid frame to place the cups under load. We converted a 20 ton hydraulic press into our test rig by adding a rigid fixture to hold the cups, and adding a pressure gage to the ram. For peace of mind, we dis-assembled the ram to measure the actual piston diameter instead of just assuming the literature it came with was correct. It was as stated, but if it were wrong, our calculations would be totally bogus.
Pressure = force/ area where pressure is in PSI, force is in Pounds, and area is in inches^2. Figuring the force applied to the cup is a simple matter of converting the pressure registered on the gage and multiplying it by the area of the ram's piston. For our tests, we loaded the cup in 200 psi increments and held the load for 30 seconds before proceeding up the scale. In most tests, a strain gage of sorts is used to measure the deflection of part at each load. The simluation from our Solidworks software showed us the most likely type of deformation, so we chose to measure that deformation using dial calipers. The good news is that the deflections we saw were pretty miniscule for the loads applied. Carbon's yield point and ultimate failure point are very close together so the lack of deflection was really of no surprise to us.
With the testing rig in house, we are now able to validate our part's strength characteristics and it will allow us to try new methods or layup schedules with new products that are surely to follow our medical centrifuge cups. And no - you dont get the numbers and no I'm not ready to divulge how we solved our silicone tooling problems- well not yet anyway! Like my friend Scott says, "a baker's success is in how he protects his recipe" Of course, if I can be of any help, I'd be glad to speak with you. Our contact info is on our website http://www.apexdesigns.net/
-Steve
Wednesday, September 22, 2010
Assume Responsibilty or Find Fault?
To Editor, Machine Design Magazine:
I've been reading your magazine now for a few years and enjoy both the technical articles and the editorials. However, this month's issue has me wielding the keyboard after reading yet another "From The Safety Files" article by contributing author and PE. Lanny Berke. This article, entitled 'Mower design flaws spark deadly fire' (with 'design flaws' in bold print) describes where a man died after being burned as a result of an accident with 30 year old lawn tractor. Certainly, I am empathetic toward the man's family and wish no one ever have to deal with that type of accident. However, I have no empathy for Berke. Here's a guy one step behind the ambulance chasing lawyers making his living at pointing out the faults and shortcomings of designers and manufacturers all day long. What I find peculiar with the story is that Mr. Berke is quick to find fault with the design of the mower, but does not offer any consideration to whether or not the mower was maintained properly or modified from its 'as delivered' condition after having been in service for THIRTY years! Mr. Berke cites that there were no warning labels or instructions about the dangers of spilled gasoline. Really? Do we need to remind Mr. Burke that his coffee is hot? Heaven forbid, there are no labels on my fountain drink cup warning me about the possible choking hazards from ice! Mr. Berke summarizes that the designers should have taken into account "reasonably foreseeable wear" and used "materials that should last a lifetime of the part". Exactly what is the expected lifetime of a lawn tractor Mr. Berke? Is 30 years beyond a reasonable life expectation for a part? Be careful Mr. Berke, your computer doesn’t have any labels on it warning you of the mail volume you’re likely to get from the angry mob of manufacturers and engineers who are fed up with a society that refuses to accept any responsibility for itself.
Steve Frank
Apex Designs
www.apexdesigns.net
I've been reading your magazine now for a few years and enjoy both the technical articles and the editorials. However, this month's issue has me wielding the keyboard after reading yet another "From The Safety Files" article by contributing author and PE. Lanny Berke. This article, entitled 'Mower design flaws spark deadly fire' (with 'design flaws' in bold print) describes where a man died after being burned as a result of an accident with 30 year old lawn tractor. Certainly, I am empathetic toward the man's family and wish no one ever have to deal with that type of accident. However, I have no empathy for Berke. Here's a guy one step behind the ambulance chasing lawyers making his living at pointing out the faults and shortcomings of designers and manufacturers all day long. What I find peculiar with the story is that Mr. Berke is quick to find fault with the design of the mower, but does not offer any consideration to whether or not the mower was maintained properly or modified from its 'as delivered' condition after having been in service for THIRTY years! Mr. Berke cites that there were no warning labels or instructions about the dangers of spilled gasoline. Really? Do we need to remind Mr. Burke that his coffee is hot? Heaven forbid, there are no labels on my fountain drink cup warning me about the possible choking hazards from ice! Mr. Berke summarizes that the designers should have taken into account "reasonably foreseeable wear" and used "materials that should last a lifetime of the part". Exactly what is the expected lifetime of a lawn tractor Mr. Berke? Is 30 years beyond a reasonable life expectation for a part? Be careful Mr. Berke, your computer doesn’t have any labels on it warning you of the mail volume you’re likely to get from the angry mob of manufacturers and engineers who are fed up with a society that refuses to accept any responsibility for itself.
Steve Frank
Apex Designs
www.apexdesigns.net
Thursday, September 16, 2010
In Motion Sept 2010
IN MOTION
The official Newsletter...uhhh Blog of
APEX DESIGNS
Up Front
Finally back in front of the keyboard, we know how much you've missed us! Its only been like 11 months since our last published 'quarterly' newsletter. We've been fortunate enough to have been pretty busy with a wide variety of projects over the last year. Unfortunately, from a news standpoint, many of these projects were proprietary in nature and we were not permitted to disclose any details. Its challenging to come up with feature topics for a newsletter when everything you've been working on is off limits. But thats not to say we've not been pecking away elsewhere. In keeping up with the trends, we've establlshed a blog and have published a few technical related articles as a way to wet our feet with blogging. Cant sleep at night? Google search under the name "apexdesignsblog" or "apexdesignsguy". If that doesnt put you to sleep, you are either a certified tech junkie or you may need to seek medical help. - Steve Frank
Relationship Building Thru R&DProviding solutions to real world problems is one of the more rewarding aspects of engineering. In August, Apex Designs was approached by LW Scientific to assist them in their quest to develop a very lightweight but extremely strong part that could withstand a force of over 2000 times its own weight. LW Scientific of Lawrenceville GA, (www.lwscientific.com) manufactures and supplies centrifuges, microscopes and related lab equipment for a variety of scientific applications. Their centrifuges are capable of rotating at several thousand rpm which can put a lot of stress on the rotating components. In an effort to decrease the cycle time of the centrifuge process, the speed at which the media must rotate has to increase. The challenge they encountered with the higher RPMs is that they were reaching the material strength limits of the rotating components. Apex Designs engineered and manufactured a carbon fiber and aluminum composite part that weighed considerably less than the same size all- aluminum part but offered considerable more strength. The pre-preg carbon parts were designed to mechanically capture a very intricately machined aluminum ring during the layup process which also added the strength of the entire part. The prototypes are undergoing comprehensive evaluation and testing and could prove to be a viable solution to allow LW Scientific to offer much greater centrifuge speeds in the future.
Sunday, August 15, 2010
Dyed Fiberglass- Cosmetic Alternative to Carbon Fiber
Some of you who know me, know that I have a particular interest in composite construction and have done alot of projects over the years both professionally at Apex Designs and personally for my racing projects. Several weeks ago, I was approached by a medical related company looking for a way to "dress up" their current machine by using carbon fiber covers in place of their existing injection molded plastic parts. After meeting with them and seeing the complexity of the parts, I quoted the project based on pre-preg oven cure carbon because the parts had several features that would be difficult to achieve by wet layup and bagging- especially since these were parts whose sole purpose was to add cosmetic appeal. My pricing for tooling and parts per unit were considerably higher than they wanted to spend for a relatively low production figures (80 parts annually) and I could not even come close to the economy of the plastic injection molded parts currently being used. Even so, the company asked me to research other alternatives to get the carbon look.
I had heard of 'black dyed fiberglass' that some folks were using to acheive the carbon fiber look and that I understood to be considerably less expensive than carbon. I researched it on the net and found that Fiberglast stocks the material as a dry fabric for about $15 per sq yard- about 1/2 the price of real carbon but still quite a bit more expensive than the Rutan glass I love to work with from Aircraft Spruce. I called Fiberglast and spoke with one of thier technical sales reps who was very helpful in sending me comparison photos of parts made with the glass and parts made with real carbon. I decided to purchase some to experiment in hopes that I could find an economical solution for my medical friends.
The glass arrived and the first thing I noticed was that it did indeed look ALOT like real carbon material- the color was nearly identical to real carbon. I also noticed that the weave was considerably tighter than standard 2x2 #284 twill weave carbon. This probably contributed to the material's lack of pliabililty- by that I mean that it was very difficult to wrap the material over even 70 degrees onto adjacent surfaces of my test tooling. Normally, my Rutan bi-directional glass literally drapes over the tool with ease and will stay put during the layup until the part is bagged. Not the case here with the black glass. I had considerable difficulty getting the fabric to stay in place over any folded edge- even when orienting the fabric at 45 degrees to the fold line- a technique that works very well with glass and carbon over tight radiused parts. I had to use ALOT more epoxy than I liked to wet out the cloth and get it to stick to the tool long enough for me to insert it in the bag and begin curing. (the epoxy used was the room temp cure West System #105 and #205 Hardener)
I let the part cure - 5 hours under vacuum, and then overnight before releasing it from the tool. With the amount of epoxy that went into the layup process, I was expecting to see a very nice smooth finish from the tooled side of the part but that was not the case here. The part had considerable voids in the corners of the weave- hundreds of tiny voids that gave the part a speckled appearance and that would be very difficult to correct even with heavy clear coating and sanding. The other interesting result was the the finished side's color was so much darker than real carbon that you could hardly even see the weave pattern- it looked more like a painted black panel.
I spent several hours clear coating, sanding and polishing the part and was never able to achieve a finish that I would ever allow coming out of my shop. I showed the real carbon sample and the black fiberglass sample to my medical friends, and they concurred with my assessment.
In conclusion, its my opinion that the black fiberglass material is not a good substitute for real carbon - especially since it's intent is primarily cosmetic. The black glass is only 1/2 the cost of real carbon, doesnt wet out very well because of its tight weave, and is too stiff a fabric to build much more than flat panels. Bottom line, if you are looking for a carbon alterative, this is not it. I would be very interested in finding people who have had success with this material, perhaps my methodology could be tuned to aid in the success of using this material.
-Steve
I had heard of 'black dyed fiberglass' that some folks were using to acheive the carbon fiber look and that I understood to be considerably less expensive than carbon. I researched it on the net and found that Fiberglast stocks the material as a dry fabric for about $15 per sq yard- about 1/2 the price of real carbon but still quite a bit more expensive than the Rutan glass I love to work with from Aircraft Spruce. I called Fiberglast and spoke with one of thier technical sales reps who was very helpful in sending me comparison photos of parts made with the glass and parts made with real carbon. I decided to purchase some to experiment in hopes that I could find an economical solution for my medical friends.
The glass arrived and the first thing I noticed was that it did indeed look ALOT like real carbon material- the color was nearly identical to real carbon. I also noticed that the weave was considerably tighter than standard 2x2 #284 twill weave carbon. This probably contributed to the material's lack of pliabililty- by that I mean that it was very difficult to wrap the material over even 70 degrees onto adjacent surfaces of my test tooling. Normally, my Rutan bi-directional glass literally drapes over the tool with ease and will stay put during the layup until the part is bagged. Not the case here with the black glass. I had considerable difficulty getting the fabric to stay in place over any folded edge- even when orienting the fabric at 45 degrees to the fold line- a technique that works very well with glass and carbon over tight radiused parts. I had to use ALOT more epoxy than I liked to wet out the cloth and get it to stick to the tool long enough for me to insert it in the bag and begin curing. (the epoxy used was the room temp cure West System #105 and #205 Hardener)
I let the part cure - 5 hours under vacuum, and then overnight before releasing it from the tool. With the amount of epoxy that went into the layup process, I was expecting to see a very nice smooth finish from the tooled side of the part but that was not the case here. The part had considerable voids in the corners of the weave- hundreds of tiny voids that gave the part a speckled appearance and that would be very difficult to correct even with heavy clear coating and sanding. The other interesting result was the the finished side's color was so much darker than real carbon that you could hardly even see the weave pattern- it looked more like a painted black panel.
I spent several hours clear coating, sanding and polishing the part and was never able to achieve a finish that I would ever allow coming out of my shop. I showed the real carbon sample and the black fiberglass sample to my medical friends, and they concurred with my assessment.
In conclusion, its my opinion that the black fiberglass material is not a good substitute for real carbon - especially since it's intent is primarily cosmetic. The black glass is only 1/2 the cost of real carbon, doesnt wet out very well because of its tight weave, and is too stiff a fabric to build much more than flat panels. Bottom line, if you are looking for a carbon alterative, this is not it. I would be very interested in finding people who have had success with this material, perhaps my methodology could be tuned to aid in the success of using this material.
-Steve
Subscribe to:
Posts (Atom)