[CncLite1]

I was considering that there seems to have been no mention of using pre-stressed materials near the edge of the "beam"... Carbon fiber for example.

------------------
Layer - High binder %, small aggregate
------------------
Layer - Carbon Fiber (mesh allowing for binder flow)
------------------
Layer - Low Binder %, Optimal packing
------------------
Layer - Carbon Fiber (mesh allowing for binder flow)
------------------
Layer - High binder %, small aggregate
------------------

Cured in a frame that allows the Carbon fiber to be put in tension during the cure.


thoughts ???

[greybeard]

Carbon fibre mesh under tension ? In one direction or both ?
It sounds like a difficult set up for the diy workshop to me, but give it a try. You may have come up with something novel.

[svenakela]

Carbon fibre is almost non stretchable. As long it's layed up in straight lines it will be close to perfection. Monodirectional fibre mats/bands are is used in everything from windsurfing masts to Formula 1-cars without pre-stretching.

I'm working with fibre laminates every day, and I'm planning to make a new table for my router with a fibre layer in the bottom and EG on top. It will be awfully strong.

[CncLite1]

simple setup ..

2 plates per end.

inside pair is held apart with 2 threaded rods.

outside plate is held off with threaded rod.

CF is clamped to outside plate.

Mold is between inner plates and associated threaded rod.

.............................

Place CF in tension with threaded rods (inside sets base tension, outside is for adjustment)

all can be done with a drill, stock plates, and a nice sized wrench

.............................


the intent would not be to stretch the CF, would be nice if possible, but to remove all "slack" from the CF to ensure it takes the tension during beam loading.

Additional materials, something "Stretchable", could be layered in to create "pre-stress"

This allows the solid matrix to be optimized to handle compressive loads.

[greybeard]

Sounds straightforward, but unfortunately my design is a little more complex than can be assembled from simple beams. I'm sure that it may be right for some people though.
John

[JerryFlyGuy]

Carbon fibre is almost non stretchable. As long it's layed up in straight lines it will be close to perfection. Monodirectional fibre mats/bands are is used in everything from windsurfing masts to Formula 1-cars without pre-stretching.

However, that being said, you will get better performance from tensioned carbon compared to hand layed carbon. Reason being is that carbon is highly suseptible to failure when it's not straight. I'm going from memory but if you use a pre-stressed carbon [pultruded rod for example] you can get as much as 30% more stress into the carbon before the fibers start to fail.. For most people it's an exotic step that is skipped, however most engineers who do composite structures will tell you that the added advantage of reduced weight when using carbon is almost negated by the fact that you have to use that much more carbon [compared to a typical 7781 glass] to reach a compareable strength level to the same 'glass'.

Basically, at the end of the day you can't use 'book values' to design something w/ Carbon unless you plan to install and place it in the manor under which the 'book values' were tested and calculated. In short.. pre-stressed layups. Not to discourage ppl from using Carbon, it looks cool.. sounds cool.. 'can' be higher performance over type 'glass' bla bla bla..

my two cents..

J

[CncLite1]

...

Basically, at the end of the day you can't use 'book values' to design something w/ Carbon unless you plan to install and place it in the manor under which the 'book values' were tested and calculated. In short.. pre-stressed layups. Not to discourage ppl from using Carbon, it looks cool.. sounds cool.. 'can' be higher performance over type 'glass' bla bla bla..

my two cents..

J

Pre-Stressed seems to intimdate people for some reason .. but if you do a little digging you will find that is a very common for a variety of reasons. Most importantly it tends to yield repeatable results.

Yes it does take some forethought ... best to consider it during the design phase such that the componets are not challanging or impossible to produce.

Regardless .. be it CF or glass or both.. I do think it is well within reach of DIY and should vastly increase the strength of EG like composites. I would hate to see someone invest all this time and end up with a big chunk of peanut brittle.

[sigma relief]

Having worked a lot with Carbon Fiber making race car parts, I would have to say that meshing it effectively with a casting mold will be very difficult. For simple structures like columns, a wet fiber layup with unidirectional carbon along the walls followed by a pour of EG would produce a decent shell, but delamination would be a large concern. I love Carbon Fiber, but I have to think judicious use of steel rods in the tension portions of the beams similar to rebar used in concrete would be much easier and more reliable. If the rods are isolated from each other, their lower vibration damping ratios will be un-noticeable. To increase binder adhesion, using coarse pitch threaded rod with spaced nuts would ensure the loads are properly transferred to the rods.

The next step in complexity again taken from concrete is post tensioning. If hollow tubes are cast into the part, cables, or more likely threaded rods can be inserted and tightened against the part. When properly done, the added compression from the bolt offsets the tension from the bending and keeps the matrix in its stronger compressive region. This post-tensioning is one of the central concepts for extending the lifetime of parts undergoing cyclical loading, while not a major concern in hobby machines, it certainly helps.

Bottom line is I just don’t think CF is worth its record high prices. Just buy more EG, no one ever said “that mill looks too heavy”

[svenakela]

Pre-Stressed seems to intimdate people for some reason .. but if you do a little digging you will find that is a very common for a variety of reasons. Most importantly it tends to yield repeatable results.

Yes it does take some forethought ... best to consider it during the design phase such that the componets are not challanging or impossible to produce.

Regardless .. be it CF or glass or both.. I do think it is well within reach of DIY and should vastly increase the strength of EG like composites. I would hate to see someone invest all this time and end up with a big chunk of peanut brittle.

My point is the same as Cnclite1 says, that the prestretching/prestressing doesn't have to be done with high forces, it's enough if the fibre is streched until it's straight. That can still be made with a hand layup, I do it often. Sail rigs (masts, booms etc) for example is rolled with higher forces, but that is to make the fibres tight and to press away resin that's not needed.
If someone wants to experiment with fibres, use monodirectional (unidirect) fibre mats. Twill fibre fabrics streth under load as the fibres are "zig zag".
Delamination is something I'm not worried about though.

In the end Sigma has a GOOD point, a mill is good when it's heavy. In my case, this time it's a router that needs a new flat table. And I have fibres. A lot of fibres.

[CncLite1]

Agreed that using CF on the surface could be problem .. aka delamination .. and complete coverage takes skill, and adds $$$

hence I recommended ..

---------------------------
------------------
Layer - High binder %, small aggregate
------------------
Layer - Carbon Fiber (mesh allowing for binder flow)
------------------
Layer - Low Binder %, Optimal packing
------------------
Layer - Carbon Fiber (mesh allowing for binder flow)
------------------
Layer - High binder %, small aggregate
------------------
----------------------------

and would add to that -- use strips with gaps between that are slightly smaller then your largest core aggregate .

11oz 1 1/2" Uni tape is ~$3 a linear yard retail .... not exactly a bank breaker on small to medium sized parts.

I agree that post tension is also a very good thing to add in ... increases overall stiffness as well as the cyclical benefits.

Weight does matter to me as I would like to use techniques are applicable to moving elements.

My motivation is to be able to provide an open source design for a 3 axis mill that can be built with hand tools and a few common power tools. Cost is a large factor as is the need to be able to move components without a hoist, forklift, etc ...

______________________
Envelope - 20" - 14" - 8".
Accuracy - .004 per Ft linear (reproducible to .0005)

1HP, 10k rpm

Very simple alignment/accurate procedures
Sealed for flood coolants

Rigid enough to mill high carbon steels
______________________


I will do some Simulations this weekend time permitting.

[brunog]

We nned to be creative in the mold making process:
- Creating a ribbed form instead of a simple bloc form.
- Insertion of styrofoam shapes to reduce overall mass can also be done.
(note that the above is already being done with regular concrete people in building construction)

Maybe CF reinforcements can be premolded to shapes that would reduce the opportunities of the EG to delaminate, making somekind of a rectangular shaped belt that would be inserted into the mold prior to pouring the EG.

Best regards

Bruno

[morphious69]

there are just to many posts here to read them all *yet* so here goes

Has anyone concedered using a vibrational dampening system similiar to the sound dampening that is used at airports? I am guessing it would require some sort of micro sensor array to capture the direction of vibrations and then an amp system that would create pulses to give "push" in the opposite direction. any ideas?

ps. if this has already been discussed then forget it and sorry about the double post of ideas

[sigma relief]

Bruno, I think your pre-fab carbon inserts would be much more reliable than co-molding wet carbon with wet EG. A CF shell is not particularly difficult, but delamination is a problem. Suspending CF inside EG is highly impractical as it will not be straight and the de-airing process will probably disturb the fibers even more. Prefab sidesteps both of these problems.

Pulltruded CF rod has predictable performance and bypasses some of the difficulty of producing your own shapes. $3 a yard for for uni tape is not bad untill you realize you will essentially be covering your machine in $1 bills several layers deep. Pulltruded rod .070 in diameter will provide similar or greater tensile strength, but with up to a 50% price premium.

I am still all for casting in hollow plastic tubes and clamping down on the structure with strong steel rods. From an enginering standpoint, it is ideal as it keeps the entire structure in compression, you simply don't have to worry about tensile failure. Adding unstressed rods, sheets, or other high tensile strength materials only serve to prevent failure by carrying a significant part of the tensile load. Unfortunately, the added material must have a very high modulus of elasticity to keep it stiff and allow it to carry the lions share of the tensile load. While All carbon fabrics and styles have a high modulus, only the priciest "high modulus" fibers provide enough of a benefit to stand apart mathmatically from cheap steel.

Unlike EG in general, the strength and stiffness improvements from adding CF rods (or any other rods) are very calculatable. By assuming a few basic parameters for the EG matrix, load sharing, equivalent stress, and equivalent stiffness can all be calculated. Unfortunately, the calculations are difficult to generalize into broad design rules, you need to run the numbers for exactly what you have in mind and then fidget with parameters one at a time to optimize. Any decent machine design or engineering analysis book will dedicate a chapter to multi material beam analysis if you are interested is the exact exuations.

John K

[sigma relief]

Morphious

Active damping systems have been deployed on industrial machinery, but are simply beyond the capabilities of even advanced DIYers. There are a variety of technical hurdles that make active damping of a solid structure much harder than 180 degree phase offset speakers in open air. First of all, sound and vibrations travel many times faster in solid materials than in low density air. This means any sensor/processor/output system needs to be much faster. Secondly, frequencies in mechanical systems tend to be higher than the anoying audible sounds muted by active cancelation. This means mass drivers (speakers) need to be designed for higher frequency with less inertia. This can make power output an issue because high frequency mass drivers can not handle high power due to their required small size.

Geometry is also an issue. Sound propegation in air is high school calculus math. Vibration analysis in machine tool structures is PhD research and expensive consulting. Software exists that gives fundamental vibration frequencies, but is not up to the level of controlling multiple reactive drivers in real time without highly specialized custom software.

Please don't mis-understand me, I'm not trying to rain on your parade. I just want convey the idea that passive systems like lots of mass and energy absorbing materials are far cheaper and more reliable than active systems.

John K

[CncLite1]

I agree 100% with Sigma on active damping, a well worded response.

Regarding the ongoing EG reinforcement topic, I think I need to provide some basic cad models to ensure concepts are communicated.

Hopefully I can steal a few hours from work this weekend. (I hate machining high carb steels and hardening to tight tolerances ... seems to take forever)

I have a Semi-DIY 7 ton press (with load cells) coming online in a week or so and should be able to do some comparative testing once it has been tweaked. Would be comparative only .. not "Standard" testing samples and methodology.

[ckelloug]

cnclite1,

I've been using ASTM D790 flexural strength methodology for 3 point bending on 7 1/2 by 1/2 x 3/8 samples. I've tested others E/G samples including jhudler and greybeard and have myself been working with the pure resins trying to formulate a better resin than the stuff we started with.

I do also have an implementation of Francois de Larrard's compressible packing model which allows me to predict the density of aggregate mixtures given the average particle size of the components.

I'm unfortunately on travel for several more weeks where wishing I was there in the lab is all I can do. I'll try to pop in here between 8:00 pm and 9:00pm for the next several weeks when I get a chance. (14 hour days on a site with no connectivity gets old).

Regards All,
Cameron

[CncLite1]

Had a few minutes to get a basic model ..

The Fiber reinforcement is used to contain the dense portion on the matrix. Ideally it is providing a minor amount of compression in the axis perpendicular to the frame.

I was going to run some FE calcs, but there are simply too many variables with potentially large variances that would make the results meaningless.

and does anyone happen to have any idea of what a "typical" vibratory power spectrum looks like for a basic cutter .. something simple like a 3/8" 2 flute at moderate load ?? (opening a can of worms here I suspect)

Edit - Found the info

http://malinc.com/publications.html

[sigma relief]

CNCLite, I am assuming the strips on top and bottom are uni-directional carbon. What is the Blue in the middle? Is it renforcenent like CF 2x2 twill or is it just representing the EG?

Just a note on the steel rails, being in the center of the part, they are on the neutral axis in bending which has no strain and thus no stress. The rods would only help when your part is in pure tension which is uncommon in machine tools. If the rods are moved to the top and bottom near the CF, they will add much more stiffness. Additionally, I don't know how many layers of Carbon you have in each stack, but if it is just one, the steel will provide many times more stiffness due to its substantiually larger area. Cheap cabon is only 50% stiffer than steel and 11 oz carbon is not very thick

Additionally, the interconnected metal structure you have will transmit vibrations much more efficiently than a non connected system such as stand alone rods, cured carbon preform in sheet or complex geometries, or hollow plastic tubes with steel clamping bolts

[CncLite1]

I should have provided a description with the image as it is misleading .. was in a rush this morning.

The steel members are for layup only and not a part of the final part.

The blue spheres are repersentative of the aggregate used in the core matrix.

The CF strips (Uni) are not intended to be a steel replacement, they are meant to increase the tensile strength of the beam via -

Containing the Core matrix (high aggregate to epoxy ratio)
Distributing localized forces to the bulk epoxy to eliminate critical failures via fracture and propogation at the surface.


------------------
Layer - High binder %, small aggregate
------------------
Layer - Carbon Fiber (mesh allowing for binder flow)
------------------
Layer - Low Binder %, Optimal packing
------------------
Layer - Carbon Fiber (mesh allowing for binder flow)
------------------
Layer - High binder %, small aggregate
------------------

[CncLite1]

I have decided to go with Ceramic preformed shapes as aggregate -

Easy to aquire
Denisty 80-90 lbs/ft^3
Available with rough surface for adhesion
Available with consistant sizing and grade
Numerous sizes and shapes

Using spheres - 2 mm, 4 mm, 5 mm, 6 mm, 9 mm - (sizes available)

Ground steel flats inserts - 3/8 x 2 x Length (.001 per ft, 4 sides +/- .001")

Alignment via Ground Granite angle plates for corners of attachment surfaces (square to .000050)

Going to use 1 1/2" 11oz Uni CF

Still trying to make a choice on epoxy. Will be using cut glass in the epoxy.


I am expecting a final density of ~100lbs/ft^3 which is higher than I would like.

Any suggestions on strategies to reduce average density and minimize any loss in strength while maintaining/enhancing dampening ?