Torque Wrench Ranges

[Steve_P]

37 lb.ft. + 120 degrees of rotation. Those are the head bolts on my Buick 3800. If the engine was completely out of the car, the angle gauge with the reaction arm might work because there would be all the room on Earth to set up the angle gauge with reaction arm.

Do you own a paint marker??? This is what I use as a DIYer. You can eyeball close enough with a paint marker to work in the real world. It's not like it's 137* +/-0.1*. And even if it was the spec, there's no way that's relevant in the real world with thread friction, head friction.... Just mark of what's close enough and you'll be fine.

Or spend $600 on a techangle to use twice in your life.

 

[dchawk81]

Do you own a paint marker??? This is what I use as a DIYer. You can eyeball close enough with a paint marker to work in the real world. It's not like it's 137* +/-0.1*. And even if it was the spec, there's no way that's relevant in the real world with thread friction, head friction.... Just mark of what's close enough and you'll be fine.

Or spend $600 on a techangle to use twice in your life.

There's a Buick for sale locally with "clickity" noises from the engine per the seller. It was probably also a 3800.

I passed because I didn't want to spend $600 for a torque angle to fix it.

 

[Steve_P]

There's a Buick for sale locally with "clickity" noises from the engine per the seller. It was probably also a 3800.

I passed because I didn't want to spend $600 for a torque angle to fix it.

 

[oldschoolcraft]

The way torque to yield works is that for the initial torque setting you tighten the bolt to something low like 25 ft-lb. The sole purpose of this initial setting is to get the slop out of the system. Next there will be an instruction to turn an additional 90 degrees or whatever. This dramatically increases the torque by as much as 75 foot-lb as you are beginning to stretch the bolt. Then finally you're told to turn an additional 90 degrees which puts you at the yield point and that gets you to over 150 ft-lb

Is there a difference between turning it 90 degrees, then turning it another 90 degrees, versus just turning it an extra 180 degrees?

Buying a 1/4 or 3/8 angle wrench is a waste of time as you will overload the wrench the first time you put it into angle mode as they tap out prior to achieving the desired angle (for a 3/8" model this is normally 100 ft-lb). Buy conventional torque wrenches in 1/4" and 3/8" and the digital with angle for the 1/2".

Great explanation! Than you! I was worried that I was going to spend $2300 on a set of three of these if I wanted to still be friends with @AEAdam . It makes a lot of sense that I'd only need the 1/2" and that also explains why the 1/2" model starts out at such a low torque rating of 15 foot pounds. Because they're not having you use it to get a 15 foot pound fastener. That would be ridiculous with a 30 inch long 6 pound tool. They're doing it because they start you with 15 foot pounds, and then have you turn it another 180 degrees maybe, and now it's ending at 150 food pounds which justifies the need for such a large tool.

 

[oldschoolcraft]

No tolerance that I'm aware of. You can only eyeball it if you can turn it in one motion. The instant you ratchet, you're out. Think about how often you get a nice clean 90 degree swing. For a lug nut you do. Not much else for me, but then again, I don't have a lift. This is the advantage of the tech angle (style) wrenches.

That sounds like what @dnschmidt was saying about turn 90 degrees, then turn 90 degrees again. I dont see the difference if it's one smooth motion or not. I'm sure there is one, I just can't visualize it and could use an explanation.

 

[garfunkle24]

If you're torquing one bolt, 2 x 90* versus 1 x 180* doesn't matter. This only comes into play when torquing a group of bolts in a sequence.

IE you go through the sequence and turn 90*, then go through the sequence again adding another 90*.

For the most part a group of bolts with a sequence is the only time you do multi-step tightening anyway.

 

[Wamsutta]

Do you own a paint marker??? This is what I use as a DIYer. You can eyeball close enough with a paint marker to work in the real world. It's not like it's 137* +/-0.1*. And even if it was the spec, there's no way that's relevant in the real world with thread friction, head friction.... Just mark of what's close enough and you'll be fine.

Or spend $600 on a techangle to use twice in your life.

It all depends on your ability to cope with paranoia and anxiety.

 

[VolvoRyan]

Do you own a paint marker??? This is what I use as a DIYer. You can eyeball close enough with a paint marker to work in the real world.

100% this.

A dab of White-Out will do as well. If you can't get 60 or 90 degree increments out of a hex faster, you shouldn't be doing this stuff.

For a lot of TTY fasteners on suspension in the rust belt, I simply buy new fasteners. Much more positive grip on fasteners that aren't half dissolved. Hardware from Volvo is dirt cheap. No joke.

-Ryan

 

[bobg03]

If you dont have a few thousand dollars in savings, which I understand more than half of the country is living paycheck to paycheck and I sympathize with you, then yes, it's reasonable to freak out about not being able to fix the car the first time. Because you might not have the money for an uber or a tow. But, there's always buses which are more inconvenient and you might have to wake up one to two hours earlier to get to work, so your PACE planning can still exist, just substitute life hassle instead of spending money.

Buses and Taxis, never been to a rural southern state, eh? Y'all know we don't get nice things like that here, we git tourists that drive like ****.

It's just a car, bruh.

You and I are both overthinkers but you are on the level that needs medication.

I have an extra glass...

 

[oldschoolcraft]

If you're torquing one bolt, 2 x 90* versus 1 x 180* doesn't matter. This only comes into play when torquing a group of bolts in a sequence.

IE you go through the sequence and turn 90*, then go through the sequence again adding another 90*.

For the most part a group of bolts with a sequence is the only time you do multi-step tightening anyway.

oh got it, like tightening lug nuts or a transmission pan. That makes sense.

 

[AEAdam]

If you're torquing one bolt, 2 x 90* versus 1 x 180* doesn't matter. This only comes into play when torquing a group of bolts in a sequence.

IE you go through the sequence and turn 90*, then go through the sequence again adding another 90*.

For the most part a group of bolts with a sequence is the only time you do multi-step tightening anyway.

I’ve never seen that with double angles. I’ve seen it with double torques.

Believe what you all want. Not trying to convince anyone or sell anything. But the Snap On models keep track of what bolt you are on. We use these at work to torque jet engines, where we could have 40 fasteners around an inlet. Break time comes around, or a shift change, the wrench verifies how many you torqued and how many are left (I think). Something like that. We have clickers in some boxes in some floor positions, but lots and lots of tech angles and we don’t specify angles (yet).

I’ve never used or needed the count functionality, but it could come in handy.

And by the way, at the end of the day these are just torque wrenches. If you use your torque wrench once a year, you probably don’t need to own one and the point is moot. When I’m working on cars, I torque every fastener that has a torque spec, (within reason) so I use these things 10x per day. They are pretty easy to use.

Snap On patented a feature they called their “early warning system“. It alerts the user when they get close. Their own internal research indicated guys tend to pull through the click or signal, over torquing. We did a similar study at work with the same result.

The fancy wrenches are very cool, but as I’ve said, not a fix for bad technique and you can do good work without one. But you need a plan, you can’t simply wing it.

 

[dchawk81]

I’ve never seen that with double angles. I’ve seen it with double torques.

Believe what you all want. Not trying to convince anyone or sell anything. But the Snap On models keep track of what bolt you are on. We use these at work to torque jet engines, where we could have 40 fasteners around an inlet. Break time comes around, or a shift change, the wrench verifies how many you torqued and how many are left (I think). Something like that. We have clickers in some boxes in some floor positions, but lots and lots of tech angles and we don’t specify angles (yet).

I’ve never used or needed the count functionality, but it could come in handy.

And by the way, at the end of the day these are just torque wrenches. If you use your torque wrench once a year, you probably don’t need to own one and the point is moot. When I’m working on cars, I torque every fastener than has a torque spec, (within reason) so I use these things 10x per day. They are pretty easy to use.

Snap On patented a feature they called their “early warning system“. It alerts the user when they get close. Their own internal research indicated guys tend to pull through the click or signal, over torquing. We did a similar study at work with the same result.

The fancy wrenches are very cool, but as I’ve said, not a fix for bad technique and you can do good work without one. But you need a plan, you can’t simply wing it.

OP isn't working on airplanes for a living.

 

[Metal-Marc]

OP isn't working on airplanes for a living.

OP is most likely a 12yo boy dreaming about tools.

 

[garfunkle24]

oh got it, like tightening lug nuts or a transmission pan. That makes sense.

Exactly, but the most common time to see a combination of a torque sequence, multi-step tightening and TTY is on head bolts.

Anything that has a tightening sequence technically should be loosened in sequence too but this is only critical in a few applications, speaking generally.

 

[dchawk81]

OP is most likely a 12yo boy dreaming about tools.

Seems to be the general consensus.

 

[Wamsutta]

OP is most likely a 12yo boy dreaming about tools.

If he's 12 years old, he's the smartest 12 year old mother effer on the internet.

I picture him in a white lab coat working in one of those computer server rooms with racks and racks of servers with wires sticking out of them.

 

[Snapped-off]

Don't forget you'll need a dial or deflecting-beam wrench to check pinion preload. Add one of those to the list too.

 

[BlitzcrankJapan]

No tolerance that I'm aware of. You can only eyeball it if you can turn it in one motion. The instant you ratchet, you're out. Think about how often you get a nice clean 90 degree swing. For a lug nut you do. Not much else for me, but then again, I don't have a lift. This is the advantage of the tech angle (style) wrenches.

Not true at all. You just look at the nut or bolt orientation and then eyeball it in. That's if you don't want to be as precise as marking it.
You can easily get it within 10 degrees with impaired sight like down the side of an engine bay. For an easily accessible nut or bolt anyone with some experience can eyeball in 3 degrees.

There is a reason pretty much all the angles are an exact 90* multiple. Because they are not really that precise. Just like torque values.
20% either way on a suspension bolt even with a torque angle sepcification will not make any difference.

 

[BlitzcrankJapan]

Don't forget you'll need a dial or deflecting-beam wrench to check pinion preload. Add one of those to the list too.

I don't believe this is true. Why do you need a deflecting beam wrench to check pinion preload?

 

[dnschmidt]

I don't believe this is true. Why do you need a deflecting beam wrench to check pinion preload?

You can also do it with a smaller digital as the torque load is shown as it adds up. This is often called trace mode.

 

[Retired dozer fixer]

OP is most likely a 12yo boy dreaming about tools.

Seems to me that if OP has to ask what size torque wrench he needs he doesn’t know what they are for or how to properly use one

 

[Wamsutta]

I don't believe this is true. Why do you need a deflecting beam wrench to check pinion preload?

Because the thing is moving while you're checking it.

 

[BlitzcrankJapan]

Because the thing is moving while you're checking it.

You mean you want to measure the force required to turn the pinion with the appropriate pinion preload?
Yes I am very aware of how many differentials are setup.

As has been mentioned already, there are other tools and ways to measure both the pinion preload and the force required to turn the pinion. This makes the quote I was replying to incorrect as you infact do not need a dial or delfecting-beam torque wrench to check pinion preload.
One of the cheapest ways is almost the same as measuring steering knuckle preload (or force to turn the knuckle). That is to use a pull scale. In the case of pinion preload which is usually specified in ft/lb you would apply the scale to a breaker bar or ratchet at 1 foot from the center of the pinion nut. The easiest way has been mentioned which is a digital torque wrench.

I would recommend nobody get a dial or deflecting beam torque wrench unless it is necessary for your work. They are virtually redundant these days.

 

[Snapped-off]

You mean you want to measure the force required to turn the pinion with the appropriate pinion preload?
Yes I am very aware of how many differentials are setup.

As has been mentioned already, there are other tools and ways to measure both the pinion preload and the force required to turn the pinion. This makes the quote I was replying to incorrect as you infact do not need a dial or delfecting-beam torque wrench to check pinion preload.
One of the cheapest ways is almost the same as measuring steering knuckle preload (or force to turn the knuckle). That is to use a pull scale. In the case of pinion preload which is usually specified in ft/lb you would apply the scale to a breaker bar or ratchet at 1 foot from the center of the pinion nut. The easiest way has been mentioned which is a digital torque wrench.

I would recommend nobody get a dial or deflecting beam torque wrench unless it is necessary for your work. They are virtually redundant these days.

What do you work on where it's specified in ft-lb?

And advocating for someone to not buy tools on GJ? Weirdo.

 

[BlitzcrankJapan]

What do you work on where it's specified in ft-lb?

And advocating for someone to not buy tools on GJ? Weirdo.

Yeah okay. You got me with the technicality. The pinion nut is set in ft lb. The force to turn the pinion would be set in inch lb.
Most of the differentials I have done are metric though. But the manuals usually specify both Nm / Ncm and ft/lb / in lb

I'm advocating for him to buy useful modern tools like a digital torque wrench that he will be able to use appropriately. Not useless ones like a dial torque wrench.

Ps. I have never used a dial torque wrench when setting up a differential or in any other bearing critical applications.
The only place I see them these days is in machine shops, old ones. I presume they just haven't got around to upgrading to modern technology yet.

P.s.s steering knuckles are often specified in kg and lbs.

 

[AEAdam]

Not true at all. You just look at the nut or bolt orientation and then eyeball it in. That's if you don't want to be as precise as marking it.
You can easily get it within 10 degrees with impaired sight like down the side of an engine bay. For an easily accessible nut or bolt anyone with some experience can eyeball in 3 degrees.

"One needs a plan" is my consistent message. Great, you want to eyeball it, someone else suggested using a mirror, don't let me hold you back. The bolts I work on get covered by cylindrical sockets, such that while I actually am turning them, I can't see their flats. You want to pull then check, great. Point is not to be snarky but make clear: "The devil is in the details".

Torque + angles are not new. But they are becoming more widespread. There are tools that address these and techniques that possibly overcome the challenge of not having the tool. To each his own.

There is a reason pretty much all the angles are an exact 90* multiple.

They aren't.

Because they are not really that precise. Just like torque values.
20% either way on a suspension bolt even with a torque angle sepcification will not make any difference.

Ok, wow. There's a sweeping statement. If torque values have a +/-20% tolerance why use a torque wrench at all?

 

[dnschmidt]

"One needs a plan" is my consistent message. Great, you want to eyeball it, someone else suggested using a mirror, don't let me hold you back. The bolts I work on get covered by cylindrical sockets, such that while I actually am turning them, I can't see their flats. You want to pull then check, great. Point is not to be snarky but make clear: "The devil is in the details".

Torque + angles are not new. But they are becoming more widespread. There are tools that address these and techniques that possibly overcome the challenge of not having the tool. To each his own.

They aren't.

Ok, wow. There's a sweeping statement. If torque values have a +/-20% tolerance why use a torque wrench at all?

Why fight stupidity? You'll never win.

 

[Hakeem]

Unless you work on trucks, a nice 20-100ftlb 3/8 torque wrench should cover the bulk of your needs. Add a 1/4” for less than 20ftlbs and you should be good.

Do you own a paint marker??? This is what I use as a DIYer. You can eyeball close enough with a paint marker to work in the real world. It's not like it's 137* +/-0.1*. And even if it was the spec, there's no way that's relevant in the real world with thread friction, head friction.... Just mark of what's close enough and you'll be fine.

Or spend $600 on a techangle to use twice in your life.

Paint marker as reference mark on the fastener, and then use your ratchet clicks to track the angle travel. Eg: 90t ratchet means each click is 4 degrees. 137deg = 34 clicks. Determine a starting point for your ratchet handle and move your ratchet back slowly, counting clicks. Pull it back to the starting point and repeat as needed until the desired angle has been reached.

This method works well when visibility of the fastener Is limited.

 

[908Jim]

You mean you want to measure the force required to turn the pinion with the appropriate pinion preload?
Yes I am very aware of how many differentials are setup.

As has been mentioned already, there are other tools and ways to measure both the pinion preload and the force required to turn the pinion. This makes the quote I was replying to incorrect as you infact do not need a dial or delfecting-beam torque wrench to check pinion preload.
One of the cheapest ways is almost the same as measuring steering knuckle preload (or force to turn the knuckle). That is to use a pull scale. In the case of pinion preload which is usually specified in ft/lb you would apply the scale to a breaker bar or ratchet at 1 foot from the center of the pinion nut. The easiest way has been mentioned which is a digital torque wrench.

I would recommend nobody get a dial or deflecting beam torque wrench unless it is necessary for your work. They are virtually redundant these days.

FWIW I disagree with this approach because you will not be able to maintain the pull scale normal to the breaker bar with any reasonable accuracy as you're turning it. You would likely start off 90 degrees but you'll measure the breakaway torque and your running torque would be observed off angle while you're constantly trying to correct the angle of the pull scale and apply a steady pull. A torque wrench eliminates most of this so reading the scale or dial accurately is your only real concern.

Dials are certainly expensive but beams are pretty affordable, easy to check against a clicker for accuracy, and very repeatable so I don't see why you wouldn't use one.

Edit: If you really want to use a pull scale for running torque, wrapping string around a socket on the pinion nut several times then pulling tangential to the socket and dividing by the moment arm (1/2 the socket OD) would give a more consistent and easily read result. For a loose socket wanting to pop off, put some electrical tape across the opening and gently tap it on to hold it in place.

 

[Hohn]

Just so we're all on the same page:

The angle requirements are all over cars now. I saw them on a 2009 VW suspension. They used to be just on head bolts. The angles are high and not easily converted into a torque "ahh, I just add 10% to the torque value to account for the angle". That won't cut it. The consequences of under (and over) torqued bolts are many and dire, especially on modern cars. I think the decision is either do it right or pay someone to do it right (hopefully). I've not seen a vehicle with angles on routine service items like lugs, oil filters, or oil plugs, etc but I suspect they are either out there or coming. Torque + Angle is a better way to preload a mechanical assembly. If you never plan on doing suspension work, maybe the advice to skip it is acceptable.

Torque + angle reduces the uncertainty in the preload. Straight torque gives you a 95% confidence band that is ±30% wide, so that's a pretty huge range. That means your M10 10.9 might be the nominal 36kN. Or it might be as low as 25.2kN. Or as high as 46.8kN. In many modern bolted joints, that level of variation is just unacceptable.

Adding the angle to the torque spec reduces the uncertain band by half. So it's ±15% instead of 30%. That's a huge improvement, one that you really need with critical joints like main caps, rod caps, head bolts, flywheel bolts.

For Rich, I bought my first Techangle in the UK. It was on eBay listed as "for parts only, not working". I recall paying about 60GBP for it, then instantly sent it out for "calibration" which cost me 80GBP incl delivery. Calibration stripped off the old handle and electronics and replaced it with new. With that wrench, I was repairing my LR Disco 3 (aka my "Chelsea Tractor" ). I think I saved a lot of money with that tool. IIRC, that vehicle had brake caliper bolt specs that were like 100nm + 270degrees. That was A LOT of torque. Not sure how I would have done that otherwise. I was working on jack stands on a cobble stone court yard. Couldn't see the bolt head from where I was squatting. Not sure how well an angle pointer thingy would have worked for me. But that's something to consider. Everyone needs a plan for how they will address this requirement. In my mind, its no different from having a code reader.

I always specify angle specs as a multiple of 60 degrees. If a tech cant figure out how to torque to a spec, then rotate X number of flats on the hex head, they probably need some remedial education. (This is one virtue of staying with traditional external hex drive screws).

That said, the angle itself need not be precisely held. Some math might illustrate why. Let's say you have a standard M10x1.5 that's 60mm grip length, something like you might see on a shock mount or such. Each rotation of the bolt is 1.5mm or just 2.5% of the bolt's grip length. Since changing the length of the fastener (stretch) is the entire mode of operation, it's pretty easy for torque+angle to surpass the precision of straight torque even if the angle amount is just estimated.

Bottom line here is that you don't need to be super precise on the angle in many instances.

As to the advice "wait and see", for me that would have been bad advice. I get pulled into car repairs with no time to prepare. That approach pretty much ensures you will either pay too much for tools or will be stuck with poor quality tools because they were all you could get in the time you had.

100% open and honest: even with 2 engineering degrees and a lifetime of working on complex things, car repairs make me a little anxious. What if something goes wrong? What if I can't get the part? What if I can't get this bolt loose? What if I strip it? What then? Will I have to face the expense of towing this vehicle to a service center? How do I get to work tomorrow? The things that give me a little confidence is knowing I have:
1) The right tools for the job. I'm not forced to make do with the wrong tool. If there's a special socket for a timing belt job for example, I want to have that upfront. "Well an Allen can be forced into a triple square and that kinda works" NOPE. Not doing that.
2) The best quality tools. My work is hard enough. Not looking to save a few bucks or even a hundred bucks if there's a chance my tools will let me down. I need to FEEL as though my tools are not going to be the limiting factor for whether this job goes okay.

Some of you will say (have said), you "fixed your Lamborghinis on the side of a dirt road with nothing more than a ball point pen and a set of craftsmen raised panel wrenches". I've actually done that (raised panel wrenches, not the Lamborghini) and it WASN'T FUN FOR ME. Fun for me is opening that pristine frictionless tool box drawer and selecting the perfect gleaming tool and having the job drama free. YMMV

Man I feel this too. Even as a "professional engineer" I'm not working on something just to work on it-- the work has something at stake. Usually that's our ability to get around (or not suffer in the process).

It's been quite the dramatic and traumatic learning curve learning this new(er) vehicle (2011 Lexus GX) and it's crazy complexity and idiosyncrasies. And as much as I have enjoyed driving the thing, I have NOT enjoyed learning the running gear and diving into it so deeply very early in the ownership experience.

 

[whateg01]

Does the use of angle reduce the need to be concerned with whether the fastener is dry or wet?

 

[Metal-Marc]

Seems to me that if OP has to ask what size torque wrench he needs he doesn’t know what they are for or how to properly use one

Nothing wrong about a kid asking about tools and cars. The next generation has to learn somewhere. Most will be happy to help and explain all the misc tools available. Just don't pretend you're an adult when it's obvious you're not.

 

[AEAdam]

Does the use of angle reduce the need to be concerned with whether the fastener is dry or wet?

Yes. The angle is independant of friction.

I highly recommend you all "work clean", especially when torquing but not only then. There are corrosion and fatigue implications to dirty components. This is one of the benefits of calling out angles for maintenance items. You don't have to be as fastidious.

 

[908Jim]

Does the use of angle reduce the need to be concerned with whether the fastener is dry or wet?

Kind of. @AEAdam knows his stuff for sure but I'll add the caveat to his explanation below that it does affect the preload achieved on the fastener with the initial torque which is the starting point for your angle based tightening. The angle measurement is a more direct relationship to preload than torque is because you remove the variability from thread fit and friction, at least in the linear range.

Yes. The angle is independant of friction.

I highly recommend you all "work clean", especially when torquing but not only then. There are corrosion and fatigue implications to dirty components. This is one of the benefits of calling out angles for maintenance items. You don't have to be as fastidious.

 

[garfunkle24]

This method works well when visibility of the fastener Is limited.

Not in any working shop I've been in. Without a stethoscope anyway.

 

[cannuck]

I very seldom use my 1/4" but I have 2 x 3/8 that cover 90% of my work on cars and 50% on light trucks. I find the 250 1/2" is great for wheel bolts, but as a clicker style if you are working down around the bottom of its range the "click" is difficult to feel or hear. My 3/4 goes all of the way to 600 (medium duty 4 link suspension) and leaves you no question you hit THAT value!! For angle I have a digital adapter as it works with breaker bar or even ratchet.

 

[BlitzcrankJapan]

So. When do I get my apology?

I was right to call out the false statement about a dial or beam torque wrench being needed, which so many of you defended.
And I was right about the torque specs not actually being that precise.

As was said already. Why fight stupidity? I'll never win......... in their eyes.

 

[Nutria]

Oldschoolcraft (OP) has a thick skin-- good on him.

I'm no torque expert, but I don't think that anyone has advised trying to stay out of the bottom 10% of the range of a torque wrench due to inaccuracy. Maybe I missed an earlier comment though. I'll use the bottom 10% in a pinch, but I can usually get into the middle range of a different wrench.

 

[AEAdam]

Oldschoolcraft (OP) has a thick skin-- good on him.

I'm no torque expert, but I don't think that anyone has advised trying to stay out of the bottom 10% of the range of a torque wrench due to inaccuracy. Maybe I missed an earlier comment though. I'll use the bottom 10% in a pinch, but I can usually get into the middle range of a different wrench.

20% is better. Post #5. But you‘ve got the right idea.

 

[Hakeem]

Some of the tasks you will want to perform in the future could involve angles. So step one, figure out how you will handle angles. Do you do that on the wrench, or with a gizmo, or with white out (marking the head)?

Next, for the sorts of vehicles most of us encounter, you will be in the 8.8 grade range. Lugs will be around 100ftlbs. But lugs will probably not be your highest torqued fasteners. Under carriage stuff can get up there. I've seen 200ftlb torques before. I think you will eventually need a 250ftlb wrench.

I do not own a 1/4" torque wrench. A digital 3/8" has the advantage that they are accurate across their entire range. Clickers are said to only be accurate from 20%-100% of their range. That would roughly eliminate their use on 13mm bolt heads.

Calibrated mechanical torque wrenches are tested for accuracy at 20%, 60%, and 100% of the maximum torque measurement value, as per ASME B107.300-2010. Assuming they are well-calibrated, a mechanical torque wrench should be accurate throughout its entire listed range, not just 20-100% of the range. Apologies if that’s what you meant and I misunderstood.

Here’s a free version of the ASME standard for anyone who’s interested in reading more:

https://www.scribd.com/document/419864822/191440347-ASME-B107-300-2010-Torque-Instruments-pdf

ISO 6789:2017 does it a bit differently, testing a TW at its minimum indicated torque value, and then at 60% and 100% of its maximum torque value.