SILCA optimal tire pressure calculator

[marciero]

This sounds like a regression fit of the data to pressures that were actually used. Whether those pressure values were fastest or in some other sense optimal is another question, especially as the evidence and the thinking and on this has evolved over the years, and some of this data is (apparently) 10 years old. I would also bet that rider preference and bias plays some role here, especially with older data.

Another thing is that this is pro data, which, as Josh points out, may not translate to average riders. In fact the performance of models like this degrade rapidly outside the range of data on which they were trained. I still it's interesting. Would be cool to see some plots or model details.

[Duende]

Appreciate the response from Josh @silca. Makes sense, and looking forward to the updates!

[simplemind]

Pressures were almost exactly what I run on gravel and road bikes. HOWEVER, it said I needed to run 10.5 PSI on my mountain bike...I don’t think so!

[Jaybee]

Originally Posted by simplemind

Pressures were almost exactly what I run on gravel and road bikes. HOWEVER, it said I needed to run 10.5 PSI on my mountain bike...I don’t think so!

I could see that for 3” tires and a light rider. The 4” tires (which are closer to 3.6”) on my fatty are best around 8-6.5psi on dirt.

[joshatsilca]

Originally Posted by simplemind

Pressures were almost exactly what I run on gravel and road bikes. HOWEVER, it said I needed to run 10.5 PSI on my mountain bike...I don’t think so!

Simplemind, what settings where you searching? Just wanting to make sure there isn't something glitchy. Also remember, the weight is rider+bike and the casing width is the measured casing width, not the tire width. Mountain tire sidewall markings show the width at the widest part of the tread and not the casing. So the 65mm casing width in the chart would be something like a 3-3.2" tire, while a tire with 2.5" (65mm) on the sidewall likely has a casing in the 55-60mm range depending on brand, style, etc.

[simplemind]

Thanks Josh, that may be the issue I had on the MTB tire. I just used the marked tire sizes on all my examples. You have to go to the manufacturers to get the casing width? And wouldn’t you also have to factor in rim (bead) width or not?

[joshatsilca]

Originally Posted by simplemind

Thanks Josh, that may be the issue I had on the MTB tire. I just used the marked tire sizes on all my examples. You have to go to the manufacturers to get the casing width? And wouldn’t you also have to factor in rim (bead) width or not?

No, we really need the casing measurement...which is a combination of the casing width from the manufacturer AND the bead width.. however, both of these things have tolerances and different manufacturers do it differently, so it isn't as easy as just saying tire size x and rim bead width Y will give you casing width Z. I've seen as much as 1mm variance in bead width between 'identical' rims and up to 2mm casing width variation between 'identical' tires.

I recommend a decent set of cheap calipers as this will remove any uncertainty, not to mention that you will find a million other uses for calipers once you have a set.

These are cheap but decent:

Analog Calipers

Digital Calipers

The ones we use in our lab

[joshatsilca]

Originally Posted by marciero

This sounds like a regression fit of the data to pressures that were actually used. Whether those pressure values were fastest or in some other sense optimal is another question, especially as the evidence and the thinking and on this has evolved over the years, and some of this data is (apparently) 10 years old. I would also bet that rider preference and bias plays some role here, especially with older data.

Another thing is that this is pro data, which, as Josh points out, may not translate to average riders. In fact the performance of models like this degrade rapidly outside the range of data on which they were trained. I still it's interesting. Would be cool to see some plots or model details.

Marciero, you are correct in how we fit the data. Working on a blog post on the method now, but essentially we used F/R balance to work a single rider's pressures into a single number, then used all of those numbers to generate the curves, then when you calculate for you, the F/R split is reapplied.

All of these data points come from pressures optimized using either Chung method or Alphamantis track measurement to determine peak minimum rolling resistance, so there is little interference from rider feel or preference in this set. I have separate overlays for rider preference, where we maybe ran a bit higher or lower than what showed as fastest for some reason or another.

Lastly, as most courses are combinations of surfaces, we typically break a race into sections and determine when the critical moves will go, where the biggest risks for flatting/equipment damage, etc are and then will develop a hybrid approach for the overall event. So we might look at fastest pressure over 4 sectors of Pave at Roubaix AND 4 of the pavement sections which we think will be critical, and then develop an overall strategy from those 8 data points. This calculator contains all 8 of those optimizations.

[weaponsgrade]

Josh, thanks for the comments. I'm looking forward to trying out the calculator. In recent years I've been using a pressure gauge and paying a lot more attention to pressures. I've found it makes a big difference - especially with mtn biking.

On a separate note, I think it's great you're continuing to stock parts for the older Silca pumps. I just replaced the plunger on my old Silca. It's a night and day difference now and I'm kicking myself for not doing it earlier.

[OldCrank]

I thought the higher CX/gravel pressures were a clincher / tubular difference thing... no?

[theboucher]

This is all very interesting. Appreciate Josh + Silca trying out some new things here!

FWIW, I have been unable to get the page to finish loading at all using Chrome, but switching to Safari makes it work fine. So, for those of you unable to access the calculator - try a different browser.

My CX pressures are the only ones I really dial in. I do MTB tires by feel/feedback and road pressure anywhere in the 70-85 ranges, depending on the road surface, etc. Interested in getting some calipers to test out some of these suggestions!