diy solar

diy solar

Cell terminal bolt torque - you might be wrong!

cinergi

1.21 Jigawatts
Joined
Aug 9, 2020
Messages
1,419
I wanted to launch a new thread because the current thinking is buried in other threads after previous values were determined to be incorrect.

TL;DR - use 4Nm max to torque your M6 bolts (or the nuts on your studs).

Explanation
Many of us were all mislead by the spec sheet of the EVE cells which says
The anti-torsion of pole is 8Nm. The torsion should be less than 8Nm when used.

We initially thought this meant the bolt torque max was 8Nm. We now believe this is the maximum rotational stress that the cell terminals can take -- in other words, don't pull on the cables that are attached to the cells.

Why are we thinking 8 is too high? Well, for one, I've stripped my threads using 7. Other have stripped at 6. So it warranted investigation ...

We found evidence of more appropriate torque values for an M6 bolt, standard pitch. There are different kinds of aluminum and we don't know which kind is used for the cells and thread depth matters, but, take a look here: - M6 is about the same as 1/4-20 and in the aluminum column we see 45.6 in/lb which is 5Nm. And here: Even SS M6 max is 60.3 in/lb. And these are values for bolts (not including thread depth)! So, even the bolts aren't supposed to go this high, let alone the aluminum.

While 4Nm sounds really low, it provides roughly 700 pounds of clamping force! https://www.engineersedge.com/calculators/torque_calc.htm
If 700 isn't enough, then your terminals/lugs/bus bars need cleaning :)

I welcome additional input/references -- but I would love to keep the thread size under control (e.g. no "me, too!" posts) so people don't get lost in here. Thanks!


References
https://diysolarforum.com/threads/x...ifepo4-cells-purchase-review.3850/post-178919 through post 2158 (and maybe beyond)
Some specific callouts: https://diysolarforum.com/threads/x...ifepo4-cells-purchase-review.3850/post-180590, https://diysolarforum.com/threads/x...lls-purchase-review.3850/page-105#post-179191

 
As reference for the non-metric peeps ... 4 Nm is 35.4 inlb.
 
I wanted to launch a new thread because the current thinking is buried in other threads after previous values were determined to be incorrect.

TL;DR - use 4Nm max to torque your M6 bolts (or the nuts on your studs).

Explanation
Many of us were all mislead by the spec sheet of the EVE cells which says


We initially thought this meant the bolt torque max was 8Nm. We now believe this is the maximum rotational stress that the cell terminals can take -- in other words, don't pull on the cables that are attached to the cells.
I would interpret EVEs spec as the maximum torque that can be applied to the terminal before the terminal begins to rotate within the cell housing.

I believe EVE sells unthreaded cells (look at the pictures in the data sheet), so their spec has nothing to do with what torque will or will not strip out threads in terminals (which were not there when they manufactured the cells).

I’m not sure whether it is EVE threading the terminals, the resellers / trading companies (many of which claim to be ‘manufacturers), or third parties such as wholesalers / middlemen, but I believe we are making a mistake to assume anything in EVE’s datasheet refers to threaded terminals.
Why are we thinking 8 is too high? Well, for one, I've stripped my threads using 7. Other have stripped at 6. So it warranted investigation ...

We found evidence of more appropriate torque values for an M6 bolt, standard pitch. There are different kinds of aluminum and we don't know which kind is used for the cells and thread depth matters, but, take a look here: - M6 is about the same as 1/4-20 and in the aluminum column we see 45.6 in/lb which is 5Nm. And here: Even SS M6 max is 60.3 in/lb. And these are values for bolts (not including thread depth)! So, even the bolts aren't supposed to go this high, let alone the aluminum.

While 4Nm sounds really low, it provides roughly 700 pounds of clamping force! https://www.engineersedge.com/calculators/torque_calc.htm
If 700 isn't enough, then your terminals/lugs/bus bars need cleaning :)

I welcome additional input/references -- but I would love to keep the thread size under control (e.g. no "me, too!" posts) so people don't get lost in here. Thanks!


References
https://diysolarforum.com/threads/x...ifepo4-cells-purchase-review.3850/post-178919 through post 2158 (and maybe beyond)
Some specific callouts: https://diysolarforum.com/threads/x...ifepo4-cells-purchase-review.3850/post-180590, https://diysolarforum.com/threads/x...lls-purchase-review.3850/page-105#post-179191


I think these materials / thread size references are far more relevant than EVEs datasheet.

So 4mm / 35 in-lb it is for me...
 
F.Y.I. If anyone is interested I found this bicycle torque wrench that covers the range of torque we should be using (3 to 10 N.m).


If that is not cool enough you can spend more than twice as much and get this click type.

 
F.Y.I. If anyone is interested I found this bicycle torque wrench that covers the range of torque we should be using (3 to 10 N.m).


If that is not cool enough you can spend more than twice as much and get this click type.

So that’s basically a bend/dial type in a housing.

3-10Nm in increments of 0.5Nm translates to 26.55 to 88.52 inch-lbs in increments of 4.4 inch-lbs.

I picked up this one: https://www.amazon.com/gp/aw/d/B07BFGN53Y/ref=sspa_mw_detail_0?ie=UTF8&psc=1

which is 0-9nm (0-80 inch-lbs) in increments of 0.25Nm (2.2 inch-lbs).
 
That looks like a good one. Cracks me up that the pointer rod is beefier than the rod that bends.
My higher-range similar design is the opposite - I suspect the pointer rod is always the same thickness while the ‘bending’ rod is various thicknesses to adjust the range...
 
That's what I ended up with too. I also got a click type but I can't hear the click. Having a visual reference is much better IMO.
Good to hear. I almost pulled the trigger on the click type but had exactly those concerns that steered me to the same style I’m used to (for auto work and PV install).
 
Good to hear. I almost pulled the trigger on the click type but had exactly those concerns that steered me to the same style I’m used to (for auto work and PV install).
No it wasn't good to hear...the click type...lol. At that low of range it's very hard to hear and I should have heeded the warnings on Amazon. I even tried a high torque setting on a different bolt and I still couldn't hear it click. It's going back.
 
No it wasn't good to hear...the click type...lol. At that low of range it's very hard to hear and I should have heeded the warnings on Amazon. I even tried a high torque setting on a different bolt and I still couldn't hear it click. It's going back.
The little one I bought just arrived. Nothing fancy but looks well-constructed (biggest complaint is that the coating protecting the markings/scale looks fragile enough that it will scratch off if not stored carefully (not the kind of tool you can just throw into the tool heat with the others.

but easy enough to zero out and seems very sensitive...
 
That's the primary reason I got a digital torque wrench. It provides two types of feedback. It vibrates and it has a nice loud beep.

Digital.
Two readings: Under-Torqued/Over-Torqued

No it wasn't good to hear...the click type...lol. At that low of range it's very hard to hear and I should have heeded the warnings on Amazon. I even tried a high torque setting on a different bolt and I still couldn't hear it click. It's going back.

You forgot to buy one of these:

1611361194314.png


I still need to get a torque wrench for lower values.
Not for battery terminals (my 3/8" wrench covers the 8mm brass screws in my AGM batteries)
Rather for breakers and the like.

Another type is kind of difficult to over-torque with: break-over, like used for SMA connectors:

4 inch-pounds, stupid price:


Several value ranges, e.g. 2.26 to 5.65 Nm

 
Double check it.

I just checked the bicycle torque wrench against my click type 1/4" torque wrench. One of them (at least) is way off.

10 N.M = 88.5 in-lb.

I set my click torque meter to 88.5 in-lb and it measured 6 N.M on the bicycle torque wrench.

Going to check my 1/4" torque wrench against my 3/8" torque wrench. If they don't agree I am going to check my 3/8" torque wrench against my 1/2" torque wrench. Hopefully two of them will agree and I can figure out which is wrong.

<time passes>

Well that was demoralizing. I am going to take all of my torque wrenches to work and check them against a stardard. We manufacture torque measuring load cells, so hopefully I will be able to see how far off my wrenches are.

And I think I am going to replace my torque wrenches. Harbor Freight, how could you have been lying to me all these years. I even got them at a 50% off sale. I paid multiple dollars ($12) for each of those wrenches.

You know the story about how the guy with only one clock is always sure what time it is? Well I have four now and none of them agree. It wouldn't surprise me to find that the bicycle wrench is the most accurate of the bunch. I did pay $30 for it.
 
I still trust the 3/8" Craftsman torque wrench my father gave me 40 years ago.

He left some cheap HF ones behind, and they have been known to strip threads in aluminum.
There's a 1/2" one which goes higher than my 3/8", suitable for some crankshaft bolts we've had to tighten. I calibrated it against the Craftsman as high as that would go, extrapolated to the torque we needed.

I'd like to set up a calibration rig that lifts weights on a lever arm.
 
I still trust the 3/8" Craftsman torque wrench my father gave me 40 years ago.

He left some cheap HF ones behind, and they have been known to strip threads in aluminum.
There's a 1/2" one which goes higher than my 3/8", suitable for some crankshaft bolts we've had to tighten. I calibrated it against the Craftsman as high as that would go, extrapolated to the torque we needed.

I'd like to set up a calibration rig that lifts weights on a lever arm.
We have several of those at work. We also have hydraulic torque calibration systems that go up way up in capacity.

The highest capacity sensor we make is 340 kN.m That is a big hunk of heat treated steel.


Here is the full range of what we do.


I designed the electronics and wrote the firmware for this one. It is used mostly in automotive dynamometers. Totally non-contact. Uses inductive power transmission and 2.4 GHz radio communication. Measures torque at up to 5000 Hz sample rates and can spin up to 10,000 RPM.

 
Last edited:
I don't really care what my wrenches say the torque is as long as I know how to correct for errors and they are repeatable. I have more confidence in the bending beam than I do in the click type to be honest. I try to use the click type near the middle of their range.
 
I still trust the 3/8" Craftsman torque wrench my father gave me 40 years ago.

He left some cheap HF ones behind, and they have been known to strip threads in aluminum.
There's a 1/2" one which goes higher than my 3/8", suitable for some crankshaft bolts we've had to tighten. I calibrated it against the Craftsman as high as that would go, extrapolated to the torque we needed.

I'd like to set up a calibration rig that lifts weights on a lever arm.
That’s how I check mine. Starting point is to measure the weight of the handle itself (unless you are using so much weight it’s contribution is in the noise...).

just holding the torque wrench horizontally in a vice (or however) from the drive end and hanging known weights (including the weight of whatever setup is used to hang them) at measured distance from the drive shaft gives a pretty reliable and cheap-and-easy reality-check...
 
Back
Top