[lgalla]

Under heat lamps the steel probably got hotter than the main E/G batch.There was more air bubbles around the steel as the heat expanded the entrapped air.One should not rely on the thick E/G mix to totally wet the steel.The steel and perhaps the entire mold should be pre wetted with neet resin first giving a better bond and perhaps a sporty finish.
Larry

[ckelloug]

Larry,

I think you are right about wetting the steel with neet resin.

yugami,

I've thought more about it and I think you and I were both half right about thermal expansion and the steel. What's going on there will be rather complicated and I still could have it wrong. I think at the deepest point of the embedment of the bar in the casting, the shrinkage of the overall block will dominate and cause a compression force pinching the steel. As one goes from the deepest layer to the shallowest layer, I believe than the shrinkage of the material towards the centers of the right and left sides will overcome the pinching effect and as one gets closer to the surface, the tendency for the composite to separate from the steel will be greater.

The extent of the effect will be related to the shear modulus of the composite and whether the epoxy actually stuck to the steel and probably numerous other difficult to quantify factors.

In summary, I think that with shrinkage, a part with a square trough for a reinforcing bar will have the sides of the trough become slightly v shaped. I'm not sure that a, this is right, and b that it is really important but I figured I'd throw it out.

--Cameron

[walter]

Thanks everyone, I appreciate the comments!

Larry's right, heat should be used for post curing only.
Surface roughening/wetting also sounds like a good idea!

Cameron,

Conrgats on the building. As always, you've got my full support.

re: drill bit
I'm not sure. Probably some kind of masonry bit

re: warping
I didn't see any of that but I think we should watch the temperature. We definitely don't want to bake the inserts.

BYK A525- I did use it and I accidentaly doubled the dosage on last 2 batches. I was in a hurry and didn't investigate but there's a chance this thing reduces surface tension and lowers viscosity.

Dang,

thanks for the kind words. It's a team effort and a lot of people contributed to the project. Check out the work of pioneers WilliamD, Sandi, Davo, Zumba, Speed- they started the whole thing.

re: bridgeport picture
Yes, it appears to be made of E/G. More of a curiosity really as it does not follow the rules.

Keep the ideas coming!
Cheers.

[harryn]

Hi Walter - nice pictures. Thanks for posting.

For some reason, I had previously missed the distinct difference in aggregate size between your mix and the commerical mix. I am sure it was beat to death in earlier posts and related to getting the mix in crevices. In any event, I like your mix esp. for surface quality.

The air pockets / density look pretty good. If you are bored some day, try making up the mix under a He purge. Even a plastic cover over the mix bucket with some welding He should be good enough. I really don't think vacuum molding will get you that much further than He. (but I am just guessing)

I have done some testing in the past bonding high power LEDs to heat sinks. These were bonded using epoxy and silver filled epoxy mixes to Al. While my test method was a bit unscientific (throwing the parts against a concrete wall as hard as possible until failure) it did clearly show the effect of cleaning the surfaces. I found that cleaning with drug store grade rubbing alcohol produced SUBSTANTIALLY stronger joints than no clean or ANY of the shop grade solvents from home depot. The shop grade solvents seem kind of oily to me.

Now I clean everything with drug store grade rubbing alcohol before bonding.

One of the other interesting effects, at least for me, was that if I traded off hard epoxy with high bond strength for slightly more flexible epoxy with less bond strength, the slightly flexible epoxy always held up to impacts much better. I have no idea if this is a general effect.

I did lightly dry sand the Al surfaces prior to bond, then did the clean.

Take care.

[jhudler]

These tests were just to see the effects of vacuum on E/G. I did not weight the mixture because I wanted to see the effects of vacuum on various compositions. In other word’s; I just wanted to play!

But first; I now officially eat crow on scoffing at using a vacuum.

I pulled 25-30 micron vacuum (25e-3 Torr) on 3 very small samples.

(all by volume)

Samples: (mixed in a small heavy Dixie (paper) drinking cup)

• 1st sample was 1 part Agsco quartz #4, #2, #2/0, 800 Zeeospheres 800, and 200 Zeeo's.
  
• 2nd sample was 1 part Agsco quartz #4, #2, #2/0, 800 Zeeospheres, and 4 parts 200 Zeeo's. (This is current formula sans #6 Agsco Brown Aluminum Oxide)
  
• 3rd sample 5 parts pool cleaner sand (cleaned in 20% HCI, rinsed, and dried), 1 part 800 Zeeo's.

How:

• Mixed in enough resin (EPOTUF 37-127 and 37-606) until it was a lump of cold peanut butter. I had to fight the urge to add more resin.
• Vibrated for 15 seconds on a palm orbital sander.
• Degassed for 1 minute after target vacuum reached.
• Repeated 4 times.

Observations:

1. Sample #1 is 1/4 inch in the cup. It inflated about 1/8 of an inch on the first degassing, and maybe a millimeter on the rest. Vibration after degassing liberated air bubbles under the surface.
2. Sample #2 is also 1/2 inch. It inflated about inch on the first degassing which then settled back to 1/8, 1/8 on the second, 1/16 on the third and forth. Vibration after degassing liberated air bubbles under the surface.
3. Sample #3 is about 5/8 inch. It inflated to almost overflow the cup about 2 inches, became very grainy and would not settle back down. Vibrating mixture after first pass recombined everything and surface gelled over again. Passes 2 thru 4 were about like sample 2.

Conclusions:

I will reiterate Walter’s statement; vibratory compaction is an absolute necessity!

1. Low frequency vibrations (30-60 Hz) appear to have more effect on the mixture than high. I was varying the speed of the sander, when it was on low the mixture stirred itself! Yep liquefaction does occur, which was a mind blower to me.
2. This is wrong because I accidently reversed the mixtures above.The aggressive inflation of sample 2 and 3, and relatively little in sample 1, may suggest that a mixture with an incorrect packing density will show this effect. This maybe a good test to determine if you have the mixture correct. Then again, did I have it correct, or did I do something different in sample 1 that caused this.

Definite Conclusions:

1. Vibratory compaction with vacuum lower than 50e-3 Torr or 50 microns is absolutely necessary to create an almost perfect void free part. I was unable to vibrate the samples while under vacuum, but I am certain this is the correct procedure.
2. Aggressive vibration during non-vacuum casting can introduce air into the mixture.
3. The perfect E/G mixing machine should be vibratory and preferably in a vacuum chamber. This will result in a much lower resin/aggregate ratio due to liquefaction.
4. It should be possible to degas a batch of E/G (vibrate and vacuum) and then place into mold while another batch is degassing. The mold should be vibrated when each batch is added, perhaps continuously.
5. 1 Minute is not long enough to degassing anything but a tiny batch without vibration and even that's iffy.

Jack (Note: it's past 12am and this might all be dog bollocks!)

Yep Bollocks it, reversed Sample 1 and 2 after looking at my notes.

[MaX-MoD]

Conclusions:

I will reiterate Walter’s statement; vibratory compaction is an absolute necessity!

1. Low frequency vibrations appear to have more effect on the mixture than high. I was varying the speed of the sander, when it was on low the mixture stirred itself! Yep liquefaction does occur, which was a mind blower to me.
2. The aggressive inflation of sample 2 and 3, and relatively little in sample 1, may suggest that a mixture with an incorrect packing density will show this effect. This maybe a good test to determine if you have the mixture correct. Then again, did I have it correct, or did I do something different in sample 1 that caused this.

Definite Conclusions:

1. Vibratory compaction with vacuum lower than 50e-3 Torr or 50 microns is absolutely necessary to create an almost perfect void free part. I was unable to vibrate the samples while under vacuum, but I am certain this is the correct procedure.
2. Aggressive vibration during non-vacuum casting can introduce air into the mixture.
3. The perfect E/G mixing machine should be vibratory and preferably in a vacuum chamber. This will result in a much lower resin/aggregate ratio due to liquefaction.
4. It should be possible to degas a batch of E/G (vibrate and vacuum) and then place into mold while another batch is degassing. The mold should be vibrated when each batch is added, perhaps continuously.
5. 1 Minute is not long enough to degassing anything but a tiny batch without vibration and even that's iffy.

Jack (Note: it's past 12am and this might all be dog bollocks!)

good job jhulder, finally someone has tested vacuum!
I have some questions thought:
-wich type of pump are you using? fridge compressor-like? 25E-3 Torr (33.25 hPa) seems very little
-approxive frequency of the vibration? Usually traditionnal concrete is degased at 10K-20KHz so I am a little confused about low frequency
I have been working with plaster and it actually liquify more et low freq than at high but degasing seems more efficient at high freq.
muste be some tricky liquid's mechanisms behind this :-s

and actually vibrating under vacuum IS what has to be done.
vacuum expands bubbles' volume. the biggest are kicked by Archimede's force alone, thought tinyest has to make their way throught the mixture (and here is where vibrations act).

Hope I'll be able to test vacuum+vibration degasing next week.

[svenakela]

Physically there's no difference between 33 hPa and 0, unless you're a atomic physician.
I have three different vacuum pumps and all of them delivers less than 10 hPa. The expensive one is measured at a lower number of molecules left than between Earth and the moon. I don't have the test certificate at the office, but I think it was 1.3E-5 or something like that.

[jhudler]

The pump I'm using is a el-cheapo eBay 2 stage pump that I got to $120 bucks. I have a filter in-line to keep from sucking aggregates in it, but this is all it will ever be used for.

I have a Vacuum Coating system for aluminizing mirrors that has a Edwards E2M5 roughing pump it gets to 10e-3, but I'm not going to risk contaminating it with degassing junk, plus it's 10 times the price. The diffusing pump will get down to 5e-9 if all the seals are fresh and expensive, and nicely chilled water for the coils.
But not for degassing! Probably never again get it clean enough to lay down a decent coating with all those polymer molecules stuck every crevice.

From everything I've read, 25 microns is great for degassing and the pumps are new and cheap. Look on eBay for seller 'txw39'.

Next I'll try and add some BYK-A-525 to the mix and see how that effects degassing in vacuum. Might help out on those big pours.

Then I need to make Cameron some test samples before he moves it off the kitchen table!

Oh almost forgot, the frequency I'm talking about is about 30-60 Hz.

Jack

[jhudler]

I had to make some corrections to my original post on Vacuum testing.
The compositions of Sample 1 and 2 were reversed.

[walter]

Great work Jack! I'm looking forward to mechanical tests.

[greybeard]

Hi all.
Just a brief post to say I'm back from great trip to Holland - train timetables that are true, and lots of cream cakes. Ho hum, back to work(???).

Following the recent postings on vacuum, I'm still rooting for centrifuging my parts

Cameron, just time to wish you the best with the exams.

Regards
John

[walter]

Good to have you back, John!

[Eson]

Could it be possible to pour the aggregates in the mold and then the epoxy ontop? I mean if you vibrate the whole thing the aggregates would begin to move and epoxy can sneak down between the space. You might think it wont mix well but i think its worth a shot.
Maybe you can add a vacuum out/inlet at the bottom of the mold so the air can disapear and epoxy is also sucked down.

[greybeard]

Could it be possible to pour the aggregates in the mold and then the epoxy ontop? I mean if you vibrate the whole thing the aggregates would begin to move and epoxy can sneak down between the space. You might think it wont mix well but i think its worth a shot.
Maybe you can add a vacuum out/inlet at the bottom of the mold so the air can disapear and epoxy is also sucked down.

I think an often overlooked problem is that if you vibrate the aggregate before the epoxy has been added, there is a strong tendency for the particles to "segregate". This will defeat the purpose of having the mixed sizes.
While I would like to have the epoxy "sneak" down amongst the aggregate, my brief experience with spinning a mix, where the G forces have reached about 50G so far, when I poured the epoxy into the middle and spun it up, it was very obvious that its viscosity still had a hard job getting it to replace the air heading for the spin axis of the tube.
Where I'd pre-mixed the E/G, the air was escaping through the mix as it was compacting. The G force being greatest at the outer surface was working to my advantage in this process.

John

[ckelloug]

Segregation is a function of aggregate design. Without digging up the De Larrard book and looking stuff up, I'll say that the mixture with the #6 aluminum oxide was designed with the formula that minimizes segregation potential. I don't know how bad the segregation would be in an actual implementation but according to the theory, variations from the formula will only make segregation more likely for the given aggregate sizes.

greybeard might be right but it may also be a very small effect. Hard data is needed.

Regards all,

Cameron

[greybeard]

Segregation is a function of aggregate design. Without digging up the De Larrard book and looking stuff up, I'll say that the mixture with the #6 aluminum oxide was designed with the formula that minimizes segregation potential. I don't know how bad the segregation would be in an actual implementation but according to the theory, variations from the formula will only make segregation more likely for the given aggregate sizes.

greybeard might be right but it may also be a very small effect. Hard data is needed.

Regards all,

Cameron

Hi Cameron. I agree that the effect may be small, but if memory serves, all of de Larrard's work used a mixing method that involved subjecting the mix to a positive pressure while the particles were being vibrated. This ensured that each time a particle could move and fill a space it did(the vibration), but it couldn't move far(the pressure) so the initial mixing was retained.
If the system was vibrated without any pressure constraint, there is nothing to stop the particles moving in a random fashion, and this inevitably leads to segregation, with the smallest particles falling, and the largest rising.

Obviously we have the presence of the epoxy to complicate matters, which de L didn't, but I think it important to keep in our minds that the recipe and methodology will interact. We don't want to loose the gain that the recipe with the measured mixture brings, by a mixing/vibrating/compaction method that might(and I stress "might") counteract it.

Again, I think it important to keep in mind that the gains that we're looking at once we move beyond a two or three component mix, are going to be proportionally smaller, and the "Law of Diminishing Returns" will almost certainly kick in.
For those like myself who are working towards a wood router set up, a two or three component(+resin) may well prove to be more than adequate.
Those looking for a metal milling machine bed, where we're pushing de L's work onwards, will almost certainly need all the extras that this amazing thread can squeeze in.

Jack, thanks for your suggestion. In fact that is the sort of pipe that I'm using. However the mold designs are made to fit the pipe, and to be spun. Bear in mind that I am producing beams, not blocks, and the spinning methods that I'm developing are aimed at an alternative methodology.
I'm starting with the same de Larrard type recipe, but with a higher resin ratio, probably nearer 20%. By packing this down into a pipe and spinning it, and I'm looking at about 150G at the outer surface, I aim to get several benefits.
1. The outward force on all the particles will be the equivalent to de L's pressure component(see above).
2. My inadequate engineering will ensure that there will be a vibration component, possibly of frightening proportion.
3. As in 1 , the force on the aggregate/resin will push the air towards the centre, (my equivalent to vacuuming out the air) along with any excess resin that is sqeezed out by the aggregate particles being denser, thus bring the resin % content down at the surface.

I'm hoping this method will remove the need for any additives, and keep the recipe as simple as possible for anyone who would not be able to obtain them.

Regards to all,
John

[jhudler]

If the system was vibrated without any pressure constraint, there is nothing to stop the particles moving in a random fashion, and this inevitably leads to segregation, with the smallest particles falling, and the largest rising.

In my testing I do not see any segregation at all. Let me scan in the cross sections of my 3 test samples and post them momentarily (err shortly).

Jack, thanks for your suggestion. In fact that is the sort of pipe that I'm using. However the mold designs are made to fit the pipe, and to be spun. Bear in mind that I am producing beams, not blocks, and the spinning methods that I'm developing are aimed at an alternative methodology.

I too am producing beams at the moment.

I do understand the desire to try alternate technologies however, you have only to purchase a vacuum pump (like those used by refrigerant techs) and supply a vibration source to start manufacturing.

What size are the beams you are producing?

Jack

[jhudler]

I think an often overlooked problem is that if you vibrate the aggregate before the epoxy has been added, there is a strong tendency for the particles to "segregate". This will defeat the purpose of having the mixed sizes.
While I would like to have the epoxy "sneak" down amongst the aggregate, my brief experience with spinning a mix, where the G forces have reached about 50G so far, when I poured the epoxy into the middle and spun it up, it was very obvious that its viscosity still had a hard job getting it to replace the air heading for the spin axis of the tube.
Where I'd pre-mixed the E/G, the air was escaping through the mix as it was compacting. The G force being greatest at the outer surface was working to my advantage in this process.

John

John,

After working with vacuum and vibration, I think I may have another way for you to achieve your goal.
I take it you have the molds for the various items you want to cast, why not try what I'm doing with the test samples. I'm using a thick walled 4 inch PVC s schedule 40 (though I don't know what you lot call it over there) capping one end and putting a 1/2 inch sheet of lexan (polycarbonate) sealed with silicone on the other.
I'm mounting it with 3 rubber bushing about 2 inches long and 1 inch diameter, then attaching the whole thing to a motor with a rod attach to an offset cam then attached to the PVC pipe.
This way I can pull a full vacuum and vibrate the mix at the same time.

I was going to do this just for mixing E/G when Gordon Bennett! that's what greybeard needs! You can find this PVC pipe from sizes all the way up to 24 inches, thought not likely to find this size down at the local DIY.

Jack

[jhudler]

Could it be possible to pour the aggregates in the mold and then the epoxy ontop? I mean if you vibrate the whole thing the aggregates would begin to move and epoxy can sneak down between the space. You might think it wont mix well but i think its worth a shot.
Maybe you can add a vacuum out/inlet at the bottom of the mold so the air can disapear and epoxy is also sucked down.

From my vacuum experments I would say this is possible but not a best practice. There would always be some void, crack, or hole somewhere that might become cloged with un-empregnated aggregate.

On the subject of infusion; Igalla is correct, it would be impossible to do this.

However, from first hand experience here's a way to mix this stuff.
First you take a.... oh what the heck I'll let the video speak for itself!

They are in Window Media format... sorry it the only format I can make.
High Res version (38 Meg)
Low Res Version (8 Meg)

150 grams of aggreate and 15 grams of resin, that's 10% right?

Jack Hudler

[lgalla]

Good stuff Jack!!!
How your mix cup didn't break,I don't know.BTW you were not wearing gloves.
As I understood the vibration process is for compacting the mix when it is in the mold.If it helps at the mix stage bonus.
Something I have always wanted;Search E-Bay for"vacuum bell jar"
Good work Jack
Larry