Mold material

General Composites Discussions Composites Talk
#1

Sorry, there’s a bit of detail and long winded.
Hi everyone! New to the forum here but have been lurking for about a week or two. I have been doing some research concerning carbon fiber since early/mid december for a project I want to do. Real work got in the way of my time but now I have some time to dedicate to this now! Many, many questions have been answered but it seems there is always something that leaves me wondering or one answer leads to a new question. So time for a question for some you guys who I would daresay are in the category of the elite when it comes to molding and composites.

Unfortunately, I hit a road block in my design until I resolve the whole process so I can design the part to be able to be molded the best.

My background: I worked about a year for a company in which my primary job was to machine large mold patterns for vehicle accessories. Engineers would send me a file, I modified the file to be able to make an accurate mold out of it, and made it on a massive CNC. The material we machined was a Wheat board MDF. We used wheat board because it leaves a fairly hard surface and we needed something that could take 300-350 degree temps during the trial period. We heated plastic, laid it on top and if the company approved the design, we made it into fiberglass molding. Basically we saturated the wood with a wax mold release and it worked perfectly fine for the most part. Now we did make good enough tooling but we never vacuum bagged anything. This was lead to some issues of course but in general things did turn out ok.

Now I want to make something out of Carbon fiber. Sure sounds easy enough I suppose as I have worked with fiberglass plenty of time, but making a higher quality is a bit different. I need to use CF for the reasons CF was made. Structural strength at the lightest possible weight. However, due the nature of what I am making, I will need to make 3-6 different molds before I have a finished design.

Items on hand:
[FONT=Calibri]- [/FONT]CNC router. Good for woods but no metals.
[FONT=Calibri]- [/FONT]An industrial vacuum pump capable of 3.75 CFM and an ultimate pressure of 7.5 X 10 -4 Torr (0.000000987 atmosphere) and a Smaller Gast pump or 2 laying around. 12.7psi is my atmospheric pressure at my altitude.
[FONT=Calibri]- [/FONT]A pressure chamber I can fit the mold in and pressurize to 150ishPSI. It is rated at 385psi but air compressor won’t go that high. It isnt technically an autoclave as it’s just a homemade tube I salvaged from an industrial heat exchanger 10.5” ID x 42” long inside. I am modifying it to have internal access to a vacuum line and return line internally, thermal elements suitable to heat to 300 degree F as well as wiring several thermal sensors I can mount on the wall, and a couple places on the mold to insure proper heating.

Question #1: What type of mold material to use? I will be machining the blank using a CNC router. Wood, foam, phenolic, etc as possible but no metals really. Paying a machinist to do aluminum is costly. Material cost about $350 for the blank and $1200-$1500 in machining costs. Phenolic (plastic composite) is nice but I cant autoclave it as it deforms when it heats up. Wheat board MDF is nice but I am concerned that a vacuum will pull the resins into the wood through any mold release essentially making it solid block of junk. Can I slightly over machine a MDF mold, say 0.100in, coat the mold in a high temp material similar to bondo but something with a better finish, and then re-machine? I know some sanding, buffing and finishing will be in order but I machine tight passes. What do you guys and gals recommend? I do have Ren material on hand but really don’t want to machine it out of this stuff. It isn’t really a good shape me anyways.

Question #2: Initially I considered just doing a wet layup with fiberglass on the 1st mold or 2 to get a feel for the spatial orientation of my design. Unfortunately, I can get close but I am not able to design a final product. 2 to 3 molds to get the general spatial dimensions correct, then fine tuning of structural point could be 1 to 2 more molds. I assume by the 6th mold a quality, useable product can be had. However, some key area of the composite will be in the 0.200 to 0.300in range or roughly 17-26 layers of 3k 2x2 5.7oz fiber. I fear during a wet layup, the resin will begin to cure before I can get the final layers on and a vacuum will not work correctly as some resin will be setting already. I estimate it could easily take 2hours to layup the first few molds, perhaps 3hrs doing wet layup. The sheer number of layers, thickness and shape dictate an infusion type is not possible but this would be a nice way to do this. Can infusion pull through 18 to 27 layers of fabric, do they make a low viscosity high temp epoxy resin?

Question 2.5: How do you lay prepreg up without getting air between the layers or will the autoclave pull the air out? I am considering prepreging my carbon fiber with some high temp stuff and then laying it up.

Question #3: If I were able to a wet layup any apply excess (not extremely excessive but liberal amounts) of an epoxy, if I vacuum bag it them and drop it into the homemade autoclave, will this remove excessive resin? Dry spots are hard to get rid of in my experience not matter how hard you try when doing a wet/painted on lay up. I know this can be an issue for determining correct resin/CF ratio though so likely wont do it.

Question #4: material. CF will be the material but in some areas the mold will be thick. 0.200 to 0.300 as mentioned above. Is a 3k 2x2 5.7oz fiber the best to use? Strength is critical. But I do have some sharp turns and the ability to drape will be very critical. I really don’t want to go smaller as the cost increases and the number of layers I need to make significantly increases. Each mold pattern nests well and I can get the patterns out of about 2.5 to 3 yards x 50” width stock material.

Question #5: Balsa reinforcement, will it suck up resin when vacuumed? How about when pressurized in an autoclave? I have a couple of small areas where I just have to have some internal structure to make a rib of some type for side support. Rib would be about 3/8” x 3/8” x 3.5” long and curved with rounding on the upward facing edge. Perhaps I could make a small 2 piece mold and inject foam into it and use that instead? What type of expansive foam would be good and not compress under autoclave pressures??

Also need a mold release recommendation for a smooth finish, how do the spray releases work? Epoxy recommendations would be appreciated as well. Mold release film type? I think I have all the rest figured out. I could figure all these things out too but there is so many types of resins out there and release waxes and everyone of them claims they are the best brand.

On materials, price matters but in truth I can spend whatever I need to but I may at some point go into production on this and need to keep my repeatable costs of consumables (Everything minus the cost of the fabric) at or under $50ish per unit. I estimate about 10oz resin per part would be used assuming 65% CF/35% resin weight. If in production, I will prepreg my own fabrics. Initially I can spend a few $k’s gearing up but I think I have the expensive pieces in place already.

I appreciate all of you who have some insight and took the time to read this!

#2

Question #1: You can mill a master in MDF with the CNC, coat the MDF with laminating resin, and heat the resin with a heat gun which will cause it to soak in. Keep on adding resin until the MDF stops soaking up resin and let it cure. If the part needs to meet high tolerances you can mill .005 over size, saturate with resin and remill. The surface will take on the feel of plastic when sanded. If you need a really good finish you can paint the MDF with a good paint and polish a perfect finish. Then you can use this positive master to pull a composite tool using high temperature epoxy resin.

Question #2: If you are doing a wet hand layup and vacuum bagging then you may have to do the many layers is stages. Layup 10 layers, vacuum bag, cure, peel the peel-ply and do the next 10. It’s not ideal but may be your only alternative unless you have help during the layup. I’ve never tried to infuse that many layers so I can’t comment.

Question 2.5: On thick prepreg layup you often need to do multiple debulking cycles through out the layup to compact the layup. Essentially part of the layup is put under vacuum but NO heat is applied to initiate the cure. Thick layups even with a high pressure autoclave can still have porosity problems in the layup isn’t consolidated well. You can also have porosity problems from any off gassing or volatile during the the B stage cure. Since you have autoclave you can apply vacuum to the chamber and vacuum to the bag so the net compaction of the layup is at 0 but the resin will still have vacuum so any volatiles can be removed. Once the B-stage is complete you can pressurized the camber to compact and consolidate the layup. Using a resin that has minimal volatiles will improve the laminate porosity.

Question #3: The vacuum bag and pressure chamber will remove the excess resin IF the breather/bleeder material is sufficient to absorb it. The chosen perf-ply between the peel-ply and breather/bleeder with help determine the objective Vf.

Question #4: Bi-axial stitched crimped free mats or multiple layers of UD material perform much better than 3K. They can also drape better as well.

Question #5: I always seal the balsa before using in a sandwich. I’ve used hairspray to lacquers. Epoxy will bond to both. You might look into some castable foams. I know they could tolerate the pressure and some might be able to handle the heat.

Since you are using higher temperatures you need to use a release that is stable at the required temperatures. Most semi-permanents will do the job. You need to test your mold surface, release, and resin combination to insure compatibility before committing. I’m a Frekote 770-nc fan but it doesn’t work with every mold surface/layup resin combination.

I didn’t answer it all but I gave it a shot. :smiley:

#3

You didn’t say what you were making and it’s none of my business but .2-.3 of carbon is awful thick. The only time I’ve ever done anything that thick is when I need to strengthen the area of a flat panel to drill a hole through it hang an engine mount. Do you need this for strength or stiffness?

You might look doing it a different way.

#4

I am looking at it many ways and you are right. But one fact I keep coming across is I need to mount 2 flanges that come together that are fairly substantial in strength and they need have bolt holes through them. These holes and the length of the structure will alternate between tension and compression. Not excessively fast but the loads will vary. I know compression is a killer to CF but I thought more about question #5 and I think I have the answer and it will solve the compression issue as well.

Stength or stiffness? Actually both. Strength is the most critical but small amounts of flex are acceptable. The CF would be post cut to relieve stress and allow slight flex to occur but evenly along its length so no region is excessively burdened. I know, or rather have researched, that machining CF will be hard on cutters and realize this will be its own bit of trial and error.

The reason for carbon fiber is to minimize weight but my theory is, a properly designed product using a mold could result in significant savings in costs. There is some other random flanges that come off that arent modeled in the attached pictures.

Its a mounting arm of sorts. Low weight is desirable. Aluminum, or magnesium could be used but it is cost prohibitive. Cast is not acceptable. Steel would techincally work but it would be a bit heavy.

I’ll add some pictures in a bit.

#5

Now this is the information I was looking for! There is some info there I have not seen mentioned anywhere as well as some good ideas. Some does lead to other questions of course but hey, thats what learning is about.

  1. Yes, the goal in the end will be to machine a positive master mold. Possibly pull several master molds off it in a production environment. I will try it just as you mentioned. If I overmill 0.005 then coat again should I coat the entire mold or is it best to let MDF ‘Breathe’ in the autoclave?

  2. Doing multiple stages would significantly simplify my process but what is the quality of bonding between a fresh layer on a cured layer? I would think even a roughed up a bit on the cured part would be the weak point when seperation would occur. Is there a way to do this without it this being the case?

2.5 - Now this is VERY interesting. I knew nothing of this. Makes a lot of sense. I will have more questions later after I have digested this concept for a while. In truth, i was not familiar with the term B-stage and found this article. Would this be a pretty accruate description? http://www.assemblyanswers.com/Assembly_Answers_2.0/B-Stage_Epoxy_Overview.html

  1. Question Answered.

  2. Perhaps a layering of Unidirectional and 3k? there will be some forces in both x and y axis. I had considered layering 3k 2x2 alternating 45 degrees each layer or every 2 layers to have stength in the x,y axis.

  3. Question Answered. You mentioned the heat, which I hadnt overly considered on foam. I have thought more on molding a rib and I think perhaps machining a corresponding mold to the rib and then casting it would be the best way. If I made it from a combination of chopped carbon fiber and thin little strips of Unidirectional fiber, it work well. I do have some compression issues I need to contend with and a solid portion here would solve it well. The weight gain would be acceptable.

And in response to your last paragraph, you are right. It should all be tested. I would just make a small test structure that is also composed of some of the other issues I am concerned about to test.

I’ll post some pictures in a few. Sigh… I get long winded when typing. I find its best to be thorough.

#6

Attached are a few pictures of a structure I hope to make. This was just a concept design for me and is not the final item as several tabs will need added and the outside shape will end up being less ‘straight’ for clearance. The tabs needed are at angles and do come up off the structure a bit which complicates the part but I can handle the corresponding molding for it. These tabs are the main reason aluminum isnt viable as the machine costs skyrocket, even if made seperate and bolt on as a seperate piece.

The design of this is split into 2 seperate layers. Well techincally they would be bonded into one component but for the structural design it can be considered to be 2 layers. Each layer split

The shape will change a bit and due to some other constraints I have, I cannot make this a hollow structure so turning it into a hollow square channel is not a possibility.

Picture #1- Arm Basic. General picture of the arm with surface edge lines removed. The wider base end on the left is bolted into place. (I have considered the possibility of galvanic corresion between SS bolts and CF- minimal but exists). The right side holds an arm. The larger hole will typically be under compession (designed for 1320lbs (600kgs) to maintain 2:1 safety margin ideally). However sometimes the larger hole may be under tension making the smaller hole under compression (designed for 440lbs (200kgs) max) but primary the smaller hole will be under tension. Holes pockets likely will have additional material laid up in this area to thicken it to distribute the stress better. As said this is just a mock up and I havent ran the stress analysis yet but CF is notoriously difficult to test with a program.

Picture #2- Arm Basic 2. Anther view of the top of the part from the other side. Shows the edge lines of the surfaces. Yes, I see the flaw in the one pocket but I already saved the pictures before I noticed it. Mostly this is for representation purposes as it isnt a final design.

Picture #3- Arm Basic 3 Underside.

Picture #4- Arm Basic 4 End. Now it gets more interesting. Ideally I would like to lay up a part using a 2 sided mold. Easier said than done but I can resolve those issues. I would like to be at a point where I have a die for the top, and bottom. Vacuum bag the the whole thing and have the mount ends pressed outward to the dies with the bag down the middle pressing them outwards. I know this can be very complex but I think it could be done. This is a main reason I expect quite a few molds revisions. Double sided molds leave good surfaces on both sides.

Picture #5- Arm Basic Cutout. Cross section of my top and bottom layers. I used 0.125 thickness per layer.

So thats the overview. Doing a dual mold and bonding them would be nice but unsure how well the bond would be.

#7

First off, that is a cool piece with lots of fun challenges. I love figuring out the manufacturing of parts like this. A few thought on fiber orientation and then I’ll give your questions a go.

The trusses between the two rails/flanges should primarily be made UD material that is oriented parallel to each truss direction. They will primarily under tension and compression. If you use a woven between the two flanges then half the material will be doing no work. Another option is to make the section between the flanges (shear web) an continous sheet of material (no lightening holes). You could make it thinner and use woven fabric (or mulitiple layers of UD material) at +/-45 to the length of flanges. UD material is going to perform best here. The top and bottom rails/flanges will also need to primarily UD material to handle the compression and tension when the structure bends . The UD material can be wrapped with wovens to hold it all together.

I’m not sure adding a core in the flanges will be all that beneficial. The flanges are carrying the bending loads and the trusses/shear web will help prevent the flanges from buckling. You could probably just use a smaller flange and no core. The flanges will be primarily under tension and compression as the structure tries to bend. The core will not be providing much of a function.

From Post #5
Question #1: I would bake the MDF after milling to drive out any excess moisture and then saturate the surface with resin while it’s still warm. That being said, the MDF wouldn’t necessary need to under high temps. You could pull a tool off the master at room temp temp, remove from master, and then post-cure the tool very carefully at elevated temps. Or a mix of both. Make the tool on the master and start to cure it a lower temps (up to 125*F), then remove the tool from the master and continue a very gradual ramped elevated temperature cute.

Question #2: When doing the laminate in stages you apply peel-ply to the last layer, then perf-ply, then breather/bleeder. Everything gets removed and the peel-ply leaves a rough surface for creating a mechanical bond. You want to polyester peel-ply not the blue teflon coated stuff. The raw polyester peel-ply will give the best bond. A chemical bond is better, but sometimes you have to do what you have to do.

Question #4: You can always orient UD material in different orientations to align the fibers with the forces.

Manufacturing Thoughts
I would highly consider a low vacuum infusion in female tools. The tools would need compressible o-rings that would seal the internal layup from the atmosphere. The tools would be loosely bolted together but enough to engage the seals. The fibers would loosely compacted in this state. Resin is then “sucked” through the tool in a vertical position with the resin feed line on the bottom and the vacuum applied to the top. Since the fibers aren’t high compacted it doesn’t take very much vacuum to infuse the fibers. It’s easy to go too fast and leave air behind. The amount of vacuum applied would dictate the speed of the infusion. 5"Hg can be sufficient. Once the part is infused the tools are fully clamped together to compress the fibers and press the excess resin out through the infusion ports. It easier said than done but I think it would be possible. This would give the you ability build the intricate layup into the tool dry with no rush.

#8

Going to digest everything you said. Had a thought about something. I’ll try and illustrate.

Your insight has been amazingly helpful. I cant stress that enough.

#9

I have been thinking about 2 possible desgin changes.

1st, I have been considering laying this up essentially as 2 procedures. I modified a cross sectionstructure to illustrate. Basically, laying up one section and curing it. Then sandblasting and/or sanding to rough up contact surface where the 2nd layer will be in contact. I also would add thin strips of CF in the layers to make a lip so the next layer would be locked into the form. Will epoxy resin ‘stick’ to cured resin when roughed up and make a satisfactory bond? Perhaps the 1st lay should have the outside piece layer of CF not fully saturated on the outside so let it remain somewhat porous?

2nd, instead of laying the part flat at the ends where seperation is likely occur, I would give it a wavy structure. The reason being is if you attempted to split 2 pieces that were flat there is no structure to hinder speratation other than the chemical bond between the even half. If you are add some structural irregularity, techincally the surface area does increase slightly but the material needs to bind on its self to come apart.

Imagine taking 2 pieces of flat sheet steel (just use something like 12"x12" in your mind) and glueing them with easy to remove material such as hot glue. You could possible take a knife, pry them a little bit apart and then get them to split. Now imagine doing the same with an equal size and thickness of corrugated steel. You might get a little seperation but pulling them apart would be extremely difficult as the steel will try and break all bonds at once since the material wont bend.

Is my mindset right?