r3volved
100 kW
liveforphysics said:
I don't, but I've queried my Facebook nerds. Maybe one of my perimeter institute friends will have access.
Conductivity improving grease project
- Thread starter liveforphysics
- Start date
I don't, but I've queried my Facebook nerds. Maybe one of my perimeter institute friends will have access.
nicobie
100 MW
wb9k said:
There is a pretty good chance that that is a flame thrower.
Dauntless
100 TW
liveforphysics said:
Darn, $31. You couldn't make do with a freebie?
http://cdn.intechopen.com/pdfs-wm/44512.pdf
Supposedly ElectronVault has succeded in using nanotech additive to grease (aka thick liquid fluid paste) to reduce contact resistance in their no weld 18650 battery packs, but they are keeping it proprietary. Rob Ferber at ElectronVault is a founder of Tesla who designed the original interconnect system for the Tzero at AC propulsion. His wife actually "invented" the nanotech grease
Their packs use foam (a cushion) to keep the pressure applied, just like Neal Saiki uses in his NTS Works no weld packs, and our own guys on the sphere do. Neal used to work for Zero Motorcycles before the money men took over. Can you show us inside pictures of the new NTS Works pack Luke? I know you have seen it. Thanks
Their packs use foam (a cushion) to keep the pressure applied, just like Neal Saiki uses in his NTS Works no weld packs, and our own guys on the sphere do. Neal used to work for Zero Motorcycles before the money men took over. Can you show us inside pictures of the new NTS Works pack Luke? I know you have seen it. Thanks
agniusm
1 MW
Just looked through a first sheet and have a question. What is the point of using grease with good dielectric properties? I thought that something like MG Chemicals 8481 would work better by increasing its conductivity by adding Tuball material?
The 'grease' is there to keep oxygen out. I think the conductivity of the graphene will be so much higher than anything else you could mix in that it wouldn't matter. I have several viscosities of the silicone grease stuff. I've used it for many years on connectors to prevent corrosion.
It might be a good feature if the mixutre is not really conductive unless you apply pressure. This way any small drips or over applications won't form a short.
My experience with solderless pack building indicates that in most cases the required pressure to make a good long term connection exceeds the crush strength of the cell ends. Hopefully this stuff will work at a lower pressure. It might be fair game to re-design the cell ends to allow for higher clamping force and be more solderless build friendly.
It might be a good feature if the mixutre is not really conductive unless you apply pressure. This way any small drips or over applications won't form a short.
My experience with solderless pack building indicates that in most cases the required pressure to make a good long term connection exceeds the crush strength of the cell ends. Hopefully this stuff will work at a lower pressure. It might be fair game to re-design the cell ends to allow for higher clamping force and be more solderless build friendly.
mistercrash
10 kW
fechter said:
Are you sure about this? I've been following the few threads about clamped 18650 packs and although I find them super interesting and try to stay positive about the concept, in the back of my mind, I'm not convinced that they would perform well for very long when considering the amount of constant vibrations our batteries go through. I'm no engineer and many "engineers" in here won't even consider the opinions of a non-engineer, but when I think of the abuse my batteries go through in a year of riding, I have a hard time conceiving that a little clamping force would give good, constant connection for an extended period.
I haven't said anything up to now, because I'm not an engineer, but I've been thinking the same thing you wrote here, that the pressure needed for a reliable connection would be too great for the cells to take it. So I've been following this grease thread also in hopes that a solution comes up.
If only there was a solvent glue to make metals bind together, like solvent glue used with Polycarbonate.
Hillhater
100 TW
flathill said:
Strong rumours on the Tesla forum that the Model 3 packs (using the 22700 cell ?) will possibly use this pressure contact "flex PCB". system instead of the " fuse wire" system of the ModS pack.
Suggestion is also that both + & -ve contact would be at the top (+ve) end of the cell , leaving the base of the cell flat for cooling/heating via a contact plate instead of the complex "snake" pipe of current packs.
Interesting concept !
agniusm
1 MW
I would not like to agree on pressure needed to make a good connection. How good connection is by spot welding? Four tiny bicy spots of good contact?
It all depends. I have ordered some pressure sensor, will use compression technique, test resistance of termination and compression .
I get that part where grease is there to keep oxygen out. what i don't understand is why use dielectric stuff and not conductive?
MG Chemicals 8481 is also grease, bleed resistant ant working temps are wider than those of ASI70:
"retains its consistency and properties over a temperature range of 50° to 200°C" for ASI70 and "Service Temperature -68 to +165 °C" for MG8481
It all depends. I have ordered some pressure sensor, will use compression technique, test resistance of termination and compression .
fechter said:
I get that part where grease is there to keep oxygen out. what i don't understand is why use dielectric stuff and not conductive?
MG Chemicals 8481 is also grease, bleed resistant ant working temps are wider than those of ASI70:
"retains its consistency and properties over a temperature range of 50° to 200°C" for ASI70 and "Service Temperature -68 to +165 °C" for MG8481
Science Faction
1 W
- Joined
- May 18, 2009
- Messages
- 59
Hi All,
Snath I like all the work you've been doing on battery packs and hopefully LFP's cnt non grease works out to improve it even further.
Excuse my ignorance with these cylindrical cells, however if you are running something simple like a 3S1P setup aligned axialy would you need more than some of the cnt paste or conductive paste between the cells if you can put pressure at each end similar to the rubber mat in Snaths pack?
How do they join/spots weld 18650 cells in axial series in things like laptop battery packs?
Thanks Paul
Snath I like all the work you've been doing on battery packs and hopefully LFP's cnt non grease works out to improve it even further.
Excuse my ignorance with these cylindrical cells, however if you are running something simple like a 3S1P setup aligned axialy would you need more than some of the cnt paste or conductive paste between the cells if you can put pressure at each end similar to the rubber mat in Snaths pack?
How do they join/spots weld 18650 cells in axial series in things like laptop battery packs?
Thanks Paul
Hillhater
100 TW
You should dismantle an old laptop pack (plenty available !) .
It's interesting to see how they arrange the cells, connect them , monitor and protect them .
If you plan on building a pack, it would be very useful for you to see.
It's interesting to see how they arrange the cells, connect them , monitor and protect them .
If you plan on building a pack, it would be very useful for you to see.
JackB
10 W
Four months later and still no data??
wb9k
10 kW
Just got around to ordering a sample of the carbon material yesterday. I have a couple of tests in mind and will share results when I have them. I've wanted to get to this sooner, but there are only so many hours in a day.....
actionobject
10 W
- Joined
- Sep 6, 2013
- Messages
- 89
Does anyone know the source of this flexible contact sheet? ?
grease lightning s.jpg
grease lightning s.jpg
wb9k
10 kW
Had to make a second request for material the other day. I used all personal info the first time and never heard anything. This time, I used the company name and address and I got a response the same day. I noticed a bit of verbiage on the website that makes me think they cannot deliver to residential addresses--"Delivery:EXW, production plant".
Hillhater
100 TW
wb9k said:
Its not uncommon for a business to be unable to supply a "retail" customer.
Many manufacturing companies just dont have the organisation , resource, or financial systems , to process a small scale retail sale.
Often it is easier for them to simply give a sample if they are convinced of the end use, ( EG, educational, charity etc).. but there is always the risk of liability if a product is miss used or causes damage later.
Usually they just avoid any risk and say no !
Voltron
1 MW
Just throwing a very unknowledgeable view in... but re pressure contact packs on bikes... maybe the bouncing actually helps the contact into being somewhat self wiping.... where if it was stationary, the corrosion build up would actually lever the contact off the surface of the cell no matter what pressure you apply.
wb9k
10 kW
Got the nano tubes today! I've only had time to run a quick and dirty cursory test with a single mechanical joint. Resistance fell from 10-12 mOhms to 1-2 mOhms--roughly a tenfold reduction in series resistance. Will be doing more tests tonight with better controls. Also will try some motor treatments with before/after shots with a thermal camera. Hoping I can get the motor to run cooler by treating the mechanical connections to the brush pigtails. This could be a big deal for high-current systems...exciting stuff!
wb9k
10 kW
Well, after a longish evening playing with this stuff and taking measurements, I'm underwhelmed and scratching my head. My first experiment measured series resistance with a pair of 6 ga. ring terminals bolted to either end of a large fuse mounted on it's standoffs. I took 3 "dry" measurements, untightening, rearranging the parts, and retightening between each measurement. Using a DC milliOhmmeter, I recorded values all in microOhms (uOhms): 515, 516, 514. Then I applied silicon dielectric grease heavily impregnated with the Tuball material to both sides of the ring terminal surface and repeated the procedure. This time I measured 514, 514, and 515 uOhms. No change. (Incidentally, the fuse itself accounts for about 470 uOhms of this total amount. Each joint between the ring terminals and the fuse terminals measured about 23 uOhms.) This was really disappointing after what I saw (or thought I saw) in the lab earlier today. I reasoned that maybe there were mechanical problems with the test setup: no surface to compress onto where the most compression is present (because of the geometry of the fuse terminals) might cause deflection that would defeat the benefit of the SWCNT's. The hardware I used earlier today was very rigid with lots of contact area and TWO bolts holding things together.
So I tried another test using a prismatic module endcap, a piece of weld strap, and a piece of flat metal hardware for making a high-power connection to a control board in a starter battery. This employs two bolts that thread into a steel backing plate. Using more or less the same procedure as the previous test, I came up with measurements of 57, 51, and 50 uOhms for the dry parts. Then I treated a large area of the weld strap surface that mated with the other piece of connecting hardware. I measured 62, 53, and 51 uOhms. Again, no real change. Sort of--I found the longer I let the treated parts sit, the lower the resistance would fall. After a couple more loosen/tighten cycles followed by some 45 minutes of settling time, the treated joint now measures 33 uOhms--about a 40% improvement over the dry parts. I seem to recall a similar pattern in the lab earlier today. I need to test settling on dry parts better, but in what I saw today they seem to stay about the same or even increase a bit with time. So, this is a bit ambiguous.
I suppose it's also possible that high current needs to pass through the circuit to make a valid measurement, using voltage drop to indirectly measure DCR. Maybe this makes a difference, though I'm not sure why it would. I should be able to test that tomorrow. I'm finding attention to details is important. The clips MUST be in the same location for every measurement. Moving a clip just an inch further down a conductor can add several uOhms to a measurement. Also, when measuring values this tiny make sure the meter is WARMED UP before you start measuring. This Instek (cheap) meter I'm using was just calibrated, but it drifted several uOhms in the first 20 minutes or so of operation. I took a whole other series of measurements that were thrown out mostly for that reason before I did the first ones I reported here. After that the meter was far more stable (you can check it periodically by clipping the leads together and verifying you measure zero.)
So, if this "settling" phenomenon doesn't turn out to be real, I hope I see something more dramatic when running real current through these joints. If neither of those pan out, I think I'll be ready to walk away. Too soon to give up just yet though.
So I tried another test using a prismatic module endcap, a piece of weld strap, and a piece of flat metal hardware for making a high-power connection to a control board in a starter battery. This employs two bolts that thread into a steel backing plate. Using more or less the same procedure as the previous test, I came up with measurements of 57, 51, and 50 uOhms for the dry parts. Then I treated a large area of the weld strap surface that mated with the other piece of connecting hardware. I measured 62, 53, and 51 uOhms. Again, no real change. Sort of--I found the longer I let the treated parts sit, the lower the resistance would fall. After a couple more loosen/tighten cycles followed by some 45 minutes of settling time, the treated joint now measures 33 uOhms--about a 40% improvement over the dry parts. I seem to recall a similar pattern in the lab earlier today. I need to test settling on dry parts better, but in what I saw today they seem to stay about the same or even increase a bit with time. So, this is a bit ambiguous.
I suppose it's also possible that high current needs to pass through the circuit to make a valid measurement, using voltage drop to indirectly measure DCR. Maybe this makes a difference, though I'm not sure why it would. I should be able to test that tomorrow. I'm finding attention to details is important. The clips MUST be in the same location for every measurement. Moving a clip just an inch further down a conductor can add several uOhms to a measurement. Also, when measuring values this tiny make sure the meter is WARMED UP before you start measuring. This Instek (cheap) meter I'm using was just calibrated, but it drifted several uOhms in the first 20 minutes or so of operation. I took a whole other series of measurements that were thrown out mostly for that reason before I did the first ones I reported here. After that the meter was far more stable (you can check it periodically by clipping the leads together and verifying you measure zero.)
So, if this "settling" phenomenon doesn't turn out to be real, I hope I see something more dramatic when running real current through these joints. If neither of those pan out, I think I'll be ready to walk away. Too soon to give up just yet though.
johnrobholmes
10 MW
A kelvin wire setup and 10 amps would be the proper method for testing it.
megacycle
100 kW
The physics of conductivity of overlapping joints is well documented.
Slipstream effect.
Conductivity is proportional to contact pressure to a point, thickness of overlap material and width of the joint, not the amount of overlap.
E.g a 30mm overlap on say a A123 tabs is no better than a 1mm overlapping joint.
With your conductive compounds, I would experiment by trying putting a.c. across them and measure volt drop, across the joint under controlled conditions, then try d.c.
Slipstream effect.
Conductivity is proportional to contact pressure to a point, thickness of overlap material and width of the joint, not the amount of overlap.
E.g a 30mm overlap on say a A123 tabs is no better than a 1mm overlapping joint.
With your conductive compounds, I would experiment by trying putting a.c. across them and measure volt drop, across the joint under controlled conditions, then try d.c.
wb9k
10 kW
I am using Kelvin probes for a 4-wire measurement, but at very low current. Thank you both for the helpful feedback.
johnrobholmes
10 MW
For repeat ability, you could solder or crimp the kelvin wires into the subject terminals. I assume you are using clips right now, as I would do at first. Higher currents are certainly easier to see changes in resistance, especially when the contact points start to heat. We do quick tests at my shop at 1 amp to sort items, and more in depth testing at 10 amps when the item will tolerate the current/heat.
liveforphysics
100 TW
Finally had time to get to doing the first tests. I took a photo of my lab book with all the raw data if someone wants to crunch it into a spreadsheet or something, it didn't seem to be better than dry, but perhaps better than just grease alone.
The biggest thing I was amazed about was how much variation occurs just from mating and unmating the connector I was using as my test fixture.
The biggest thing I was amazed about was how much variation occurs just from mating and unmating the connector I was using as my test fixture.
