Mineral oil in freezing conditions

Dunno about the specific cause, but I've only experienced it once -  in about minus-10, up Skiddaw in the snow.

My Shimano brakes have been fine in the cold snaps down to about minus-5  every year since.

How cold were you talking about?

About -5, I reckon.  This is my commuter bike so I keep it in the shed.   It's therefore had plenty of time and freeze up.

I've had this a couple of times on my Cube Fatbike. It only seems to affect my front brake and feels like it's a bit "wooden". This would be after riding for a few hours in minus 5C or so, likely in conditions that I'm not braking often/heavily and therefore not getting any heat into the system. Your theory seems reasonable and I wonder if a simple top-down bleed might resolve it. Certainly worth trying for little effort.

You'll possibly have more issues with frozen freehubs for a commuter stored outside.  Mine is in the garage and the only issue I've had this year is the rear mech cable froze after I got the bike out. Took a few miles of attempting gear changes for it to free up.

I've had frozen brake cables, freehub freezing and gears. In prolonged periods of cold, I used to bring the bike in overnight, so it didn't freeze up. I don't currently have that option, but the garage has a dehumidifier on, which warms it up just enough.

Your theory seems reasonable and I wonder if a simple top-down bleed might resolve it. Certainly worth trying for little effort.

That's the plan for the weekend. I'll probably take the calipers off and try to move them around a bit while the fluid is running through just in case some water has pooled somewhere that for some reason a bleed can't get to.

From Monday onwards the forecast isn't showing that the temperatures are going to be as low as today so not sure if that's going to tell me much.

This will let you calculate how much thermal expansion you'll be getting in that small blob of water ... https://www.engineeringtoolbox.com/volumetric-temperature-expansion-d_315.html It isn't much - could it be some ice forming, a change in viscosity?

I didn’t see a definitive answer but if I had to take a guess I would assume it’s down to the hydrophobic nature of mineral oil.

By that I mean that, if any water does get into the system, it will most likely ‘pool’ somewhere at the bottom of the caliper. You then end up with that pocket of water freezing and expanding which causes the effects everyone seems to be describing.

Obvious way to test the theory would be to check it early in the morning, then pour hot water over the calipers and hoses to melt the supposed ice.

I'm skeptical that ice expanding is the problem. Unless the system is overfilled, the ice will just force fluid back into the reservoir and the brake lever should be at the same level.

Obvious way to test the theory would be to check it early in the morning, then pour hot water over the calipers and hoses to melt the supposed ice.

I've considered stopping mid-ride and pissing on my front brake but finding my willie when it's minus 5C is a whole different problem 😂

Looking at the design (and assuming any water is going to migrate to the lowest point) it wouldn't be surprising if you could actually end up with quite a lot of water right next to the calipers that simply couldn't be shifted with a gravity bleed.

https://bike.shimano.com/en-EU/technologies/component/details/one-way-bleeding.html

By that I mean that, if any water does get into the system, it will most likely ‘pool’ somewhere at the bottom of the caliper. You then end up with that pocket of water freezing and expanding which causes the effects everyone seems to be describing.

This has to be a thing. How much of a thing I don't know, but that same water vapour is also a problem with hot brakes and brake fade.

Don't forget that seals around the pistons and master cylinder, plus the oil itself will stiffer and more viscous.

How do people think all this water gets into a sealed brake system?

Its the same as with forks, shocks
and dropper posts. Incrementally past sliding seals. Why else do you think fluid needs changing regularly?

Forks and shocks cycle hundreds of times per hour. The seals in brakes cycle about once per few months (i.e. every time you change the pads). If the seals were so leaky that they let water in in wet weather, you'd be constantly having to top up brake fluid in dry weather because fluid would be leaking out (the pressure is on the inside, not the outside). The only real mechanism for significant amounts of water getting in is if you blast the brakes with a pressure washer.

On top of that, even if the fluid in the caliper expands (like it does when it gets hot), the system is designed so that the excess fluid flows back to the reservoir and the bite point remains constant.

I'm not saying that there isn't a change in the bite point, just that ice expanding in the caliper doesn't seem a likely explanation.

Its the same as with forks, shocks and dropper posts. Incrementally past sliding seals.

No, it's not - the seals in brakes flex, they don't slide, and that only happens when the system is under pressure which means oil comes out, water does not go in.  The difference with DOT fluid is that the fluid itself absorbs water on a molecular level so it can suck it in through seals, theoretically.  I'm even sceptical about this since unlike a car, the master cylinder is also sealed from the atmosphere.

Why else do you think fluid needs changing regularly?

I don't think it does for mineral oil.  If it does, it's probably related to the oil breaking down or changing characteristics over time, with heat.

the seals in brakes flex, they don’t slide, and that only happens when the system is under pressure which means oil comes out, water does not go in.

They do slide very gradually as the pads wear. Each brake pad change is basically equivalent to one stroke on a fork or shock. If you wear your brake pads out in six months and replace them, that's the same as doing one big jump, they're both one cycle of the piston and seals. It's just that you cycle the suspension hundreds or thousands of times per ride, but you only cycle the brakes once every few months.

So if we've excluded water contamination, whats the cause?

How old is the bike? I'd be surprised if enough water has built up to cause a problem after several winters

Anyone who has been involved with testing hydraulic systems will likely have seen that it's more often the low pressure test that causes problems rather than the high pressure test.

It's very difficult to design a hydraulic seal for a flexible hose that it going to work with both high and low pressure (and possibly even sometimes a slight negative pressure depending on the movement).  Obviously with a brake the first priority is to create a high pressure seal and the low pressure seal is secondary even though it's the state the bike spends the vast majority of it's time in.

Also, it's difficult to design a seal that works perfectly in both directions.  Again, the priority is going to be towards keeping fluid in.

The idea that fluid or air could be drawn into a brake system is not that far fetched and the fact all brake manufacturers tell you to bleed your brakes regularly suggests it's expected behaviour.

Looking at the picture I posted above, the first type seems like a really poor design choice for a brake that is going to use mineral oil.  The two chambers are acting as sumps for water and no amount of bleeding is going to shift that water once it gets in there.

Short term I'm going to modify my shimano gravity bleeds so that I take the caliper off and orientate it so that the bleed nipple is the lowest point of the system.  Hopefully that will cause any water droplets to sink end up in the waste bottle

Long term the solution is just to stick to brakes that use DOT fluid.  At least their performance is going to degrade gradually rather than catastrophically (more thinking about overheating here rather than freezing).

Also, it’s difficult to design a seal that works perfectly in both directions. Again, the priority is going to be towards keeping fluid in.

The idea that fluid or air could be drawn into a brake system is not that far fetched and the fact all brake manufacturers tell you to bleed your brakes regularly suggests it’s expected behaviour.

The piston seals at the calipers are rectangular section o-rings. They will be equally as effective at keeping water out as in keeping brake fluid in. The seals in the master cylinder are different and pressure washing them could blast water into the system. It's possible that driving in rain with the bike on a roof rack might too. Riding the bike normally in wet weather shouldn't be a problem though.

The main problem with the theory that ice expansion is causing the lever bite point to change is that brake systems are designed to allow for fluid expansion. There's a transfer port between the reservoir and master cylinder. If there was expansion due to ice formation, it would just force fluid back into the reservoir and the bite point should stay the same.

An exception to that would be if the system was underfilled. The way to test that would be to make sure the reservoir is filled and then see if the bite point changes. If it does, then you need to readjust the lever and see if the changing bite point continues when the bike is left out in freezing conditions.

The piston seals at the calipers are rectangular section o-rings. They will be equally as effective at keeping water out as in keeping brake fluid in. The seals in the master cylinder are different and pressure washing them could blast water into the system. It’s possible that driving in rain with the bike on a roof rack might too. Riding the bike normally in wet weather shouldn’t be a problem though.

If you're saying that bike brakes should, in theory, never need bled then I think the experience of most riders would disagree with that.  However it happens, air sometimes gets into the system so it's not too much of a stretch to imagine water might also get in.

The main problem with the theory that ice expansion is causing the lever bite point to change is that brake systems are designed to allow for fluid expansion. There’s a transfer port between the reservoir and master cylinder. If there was expansion due to ice formation, it would just force fluid back into the reservoir and the bite point should stay the same.

Not necessarily.

If there was a small pool of water at the bottom of the caliper and it froze you would expect it to freeze from the top, thus sealing it off from the mineral fluid in the rest of the system.  Then it will just take the path of least resistance which will be to push the calipers out.

I've used various Shimano brakes for years now and am very familiar with aub-zero temperature woodenness. I've had it with virtually new brakes, recently bled ones and old, neglected ones too. I've always assumed that it's something to do with the seals in the caliper stiffening up when they get really cold, so the piston doesn't retract as quickly after use and the brake lever gradually pumps up as a result.

I'm not an engineer or a materials scientist, physicist or whatever, but that's what I've always thought was the most likely explanation and, in true forum style, I am now 100 per-cent certain this is the case 😉

Anyway, it's a theory.

ps: what we need is for someone with the aftermarket Hope calipers fitted to a Shimano road hydraulic system to use their brakes back-to-back with a Shimano caliper system, both with used mineral oll, in really low temperatures to see if they behave differently. Or, I guess, identical Shimano systems with new and old mineral oil, but as above, I've had the cold weather issue with pretty much new brakes before now.

the fact all brake manufacturers tell you to bleed your brakes regularly suggests it’s expected behaviour.

I do wonder what their reasoning is for this. I'd also be keen to know (genuinely not sarcastically) if you find any water in there when you try the inverted bleed.

FWIW I almost never bleed my brakes and they always work spot on. I did find some corrosion in the seal bed of my Mono Minis when I changed the seals after about a decade of use so that suggests some water had got into the DOT.

I’ve used various Shimano brakes for years now and am very familiar with aub-zero temperature woodenness.

Did any of them have the 'One Way Bleeding' feature (see the picture above) or did they all the bleed port at the top?

Did any of them have the ‘One Way Bleeding’ feature (see the picture above) or did they all the bleed port at the top?

No idea. I just assume it's going to happen with all Shimano brakes and am rarely disappointed. It's irritating in sub-zero conditions, but thanks to Shimano's invention of Climate Change™ I generally forget that it ever happened until the next time.

Edit: * goes into cellar, looks at assorted bikes * - they all have bleed nipples at the same end as the hose attachment. Which Shimano brakes are different then?

It could also be a QA issue, of course.  With moisture somehow being introduced during assembly.  Shouldn't happen but then a lot of things shouldn't happen.

No idea. I just assume it’s going to happen with all Shimano brakes and am rarely disappointed.

Fair enough.  You could try taking the caliper off and do a gravity bleed making sure the bleed nipple is at the lowest point and see if it happens again.  If you remember 😉

I do wonder what their reasoning is for this. I’d also be keen to know (genuinely not sarcastically) if you find any water in there when you try the inverted bleed.

I suspect the chamber behind the caliper isn't as big as it looks in the picture so I reckon seeing anything is going to be tricky.  I'll keep an eye out next bleed though.

Edit: * goes into cellar, looks at assorted bikes * – they all have bleed nipples at the same end as the hose attachment. Which Shimano brakes are different then?

Those would be the non-One Way Bleed feature.

I do wonder what their reasoning is for this.

I can only assume it's to take into account poor manufacture. The tolerances of these mass-produced items are probs not really good enough so that [for instance] seals sometimes don't seal quite as well as they should, or as parts wear, air can creep past them.

FWIW I almost never bleed my brakes and they always work spot on.

Likewise, I will try to leave them well alone. I'm generally one for preventative maintenance on bikes, but brakes are left unless they're actually not working as intended.

Like BWD I always assumed the wooden feeling in Shimano brakes was something to do with either 1. low temps effecting mineral oil viscosity, or 2. some property of the material they made the seals out of. I never thought it might be water trapped anywhere though...

It could also be a QA issue, of course.

TBH, on all of my bikes over the years, all with Shimano brakes, I've only experienced the problem on one bike and only on one of the brakes. It does make me wonder what the variable is. Inevitably, it does tend to have been the fatbikes that have been exposed to extreme cold for the longest periods, but not exclusively so. I've camped out overnight in temperatures below -5C and not had an issue with my brakes the following day.

If there was a small pool of water at the bottom of the caliper and it froze you would expect it to freeze from the top, thus sealing it off from the mineral fluid in the rest of the system. Then it will just take the path of least resistance which will be to push the calipers out.

That would cause the brakes to jam on.

Sub-zero conditions will affect the oil viscosity and probably the seals. I wouldn't be surprised if it made them feel slightly different, but the ice in the calipers theory just doesn't seem very plausible.

Another possibility is that, if there is air in the system, the air bubbles might shrink in the cold so the brakes would be less spongy.

Another possibility is that, if there is air in the system, the air bubbles might shrink in the cold so the brakes would be less spongy.

I'm pretty sure it's not that. I used to run my old Shimanos "over-filled" to achieve the feeling/bite point I liked and my cold-weather experience feels nothing like that.

I'd previously wondered if it was a build-up of ice external to the hydraulic system - perhaps around the pad itself. I should really get out today and do some testing - it's currently -8C - but I'm waiting for a gas engineer to turn up as we've no central heating!

It may be possible that the pads absorb moisture and then that freezes.

That would cause the brakes to jam on.

That's what happened to me yesterday.  Well, they didn't jam completely on but the pads were in contact with the discs and stayed that way all the way to work.

Keeping it in the above freezing garage for 8 hours got the pads to become free from the disc but still with far less lever travel.

It really feels like the pads moved into contact with the disc and the fluid from the reservoir moved into the 'active' part of the system causing them to stay there even after the brakes were kept in above freezing temperatures, just now with the pads at least able to move back from the disc.  Sort of like how people do their own adjustments to the throw of the lever.

I live in a Nordic country and regularly cycle in winter conditions, I switched to this mineral oil a year ago on my Shimano brakes -

https://www.nils.eu/en/prodotti/clutch-command/

This is a compatible mineral oil that will not change in feel as much as the Shimano mineral oil - my brakes feel normal when it is down to -10 c. Colder than that and the brakes start to change, but they are still much better than with original Shimano oil.

With official oil brakes go really weird in temps lower than -10, and lever throw decreases a lot down to almost nothing before biting.

Any water in the system is most likely to enter via the lever.

The external master cylinder seal only needs to seal between the reservoir and the environment, and it'll mainly be designed to keep fluid in, not water out.

I can't be bothered searching for Shimano lever exploded diagrams, but looking at a couple Hope's, current levers are relying on what looks like a standard O-ring, whereas earlier levers were using a cone seal set to provide maximum fluid retention.

If that seal is surrounded by water, then chances are every time you operate the brake, there is a risk of water passing it. Combine that with heat expansion/contraction, and you also have the possibility of water being drawn in as well.

Once in, the water will settle to the bottom of the system. If enough of it was to gather in the caliper, there is theoretically the possibility that it could freeze, expand, and push the piston out.

However you also need to consider other issues like the seal materials. One reason DOT fluid is used, is it is compatible with natural rubber. Natural rubber retains it's flexibility in sub-zero conditions, whereas other materials start to noticeably stiffen up.

So it could well be a combination of things causing the issue.

This is a compatible mineral oil that will not change in feel as much as the Shimano mineral oil – my brakes feel normal when it is down to -10 c. Colder than that and the brakes start to change, but they are still much better than with original Shimano oil.

Is the theory then that Shimano mineral oil increases in viscosity in low temperatures, but this stuff doesn't or at least not to the same extent and that's what makes the difference?

Is the theory then that Shimano mineral oil increases in viscosity in low temperatures, but this stuff doesn’t or at least not to the same extent and that’s what makes the difference?

Yes.