What needs to be corrected ?

There might even be a hookless disc for 28 mm tires on the horizon with their entire lineup converting.

let's stipulate that the chart you show reflects accurately the Crr on existing surfaces. that's a big assumption, because you have al morrison data, tom anhalt data, drum testing, field testing, all showing up on the same charts. still, these are smart folks (al, tom, josh) so let's just assume all of that is more or less accurate.

i don't see that in the data i presented in my article that zipp discusses variable road surfaces. perhaps i just still don't get your point. show me in the article i wrote where zipp is arguing road surface texture was even broached. for sure, the morrison test showed a higher optimized pressure. but let's stipulate it's accurate. that was a 23mm tire, tubed, on morrison's drum. it was not a 25mm or 28mm tire, tubeless, on a different surface using a different test rig paradigm.

all the testing zipp did was on one surface, that zipp considered representative of a typical event (triathlon or road race). what i would say - and did say - is that ideal pressures on zipp's chart seem curiously low for the 858 NSW. as i noted above, i'd like to know why that is. but we've come a long way since 100psi was the optimal pressure. we're not running 23mm tires anymore, many of us have moved past 25mm tires, and the pro peloton in europe went from completely-anti tubeless to "you must provide us tubeless!" in 1 year (from last year to this).

i would make these points:

1. did you ride 28s in your real world testing? it might be my imagination, but i get the impression that the things-that-roll cogniscenti are biased against that tire size, just as they were at one time (not that long ago) biased against 25mm, wider inner bead widths, disc brakes, clincher, then tubeless, then hookless. (but they weren't dumb; they were very obviously right about one thing: lower pressures.)

2. throw off any bias you might hold against 28mm tires, and embrace the possibility that this tire width in concert with tubeless, hookless, new well designs might conspire to create a system that reacts differently than what you're used to.

3. your say that your real world testing backs up drum testing, as i understand it. zipp says its field testing validates its rider-on rolling road Crr testing. i don't where that leaves us, except with more testing to do.

as i said in the beginning yes, there are some things that don't make sense to me, that need explanation, and i listed what those were. you think those inconsistencies make this an uncompelling wheel for kona. i find the data compelling, and believe the wheel ideal for kona. what accounts for the space between us?

i recognize both the apparent inconsistencies in rolling road testing and in drum testing, and i try to make sense of them both; and glean what's of value out of each. the experts minimize, gloss over, or reject the inconsistencies in drum testing. i look forward to the conversation we might have when we throw it all out there on the table and deal with the data we like and the data we don't like.

---

Dan Empfield
aka Slowman

of this i'm sure: if there isn't now there will be. i'm not 100 percent on this, but i think perhaps blu and iden might be running a hookless CADEX disc. i have to run that down.

---

Dan Empfield
aka Slowman

I don't see how iden could run a large hookless tire on wide internal rim when the trinity limits (according to Giant) are 25mm tires with 28mm external rim...

---

https://besse.info/
https://www.strava.com/athletes/2012033

i'm not sure. i don't know. i guess i'm more thinking of what blu is running, because he can stick a motocross tire in that CADEX frame of his. i assume iden's running what blue is running but maybe not. also, i might have gotten bad info on what blu is running. that's why i've got inquiries on on that.

---

Dan Empfield
aka Slowman

The annotation of the y axis on these graphs is wrong for years. I can‘t really trust such a proposition when a simple error isn‘t recognized for years.

In this way these ZIPP findings have the smell of marketing noise. What are the error bars with this rolling road? How are the test conditions not been chosen just to verify the own marketing claims? Are high Crr on these set up due to set up specific resonance effects?

Let me try this another way. When I read the article I look at the red line (old wheels, 25mm tires). This should be my reference, something I can relate to.

Then I look at the improvements with the new system.
My reference is the red line and the improvements are in the yellow/orange lines

The problem is the red line does not compute for me. If I cannot wrap my head around the "reference" I cannot appreciate the possible improvement

The reason the red line does not compute is that it doesn't match data I have seen from Tom, Silca or my own experience. So I try to understand it.

Then I realized the trend Zipp was showing matched the trend Silca showed for bad road surfaces. So maybe (I don't know) their data is based bad/rough surfaces. I don't know, but it's the only way I can make sense of the trend they see for the red line (my reference)

Maybe their data is valid for Paris Roubaix, but not my local formula 1 track.

I would be quite easy for them to just show a picture of the 'rolling road' you'd think?

---

https://besse.info/
https://www.strava.com/athletes/2012033

Now I apologise if what I say is completely obvious and I've missed something in the previous posts, but I just wanted to check. There's a lot of really complex discussion here, but there's a 'simple' first stage.

A lot of the comments are about why there's not a difference in the red line as pressure changes. The answer is that this is a delta plot and that red line is the baseline of zero. So there are differences, it's just that line is set to be zero with the other lines based off that.

Where this entire chart falls down (to me) is that I can't see where the baselines are able to be set for portions of the graph that don't overlap. Which is where I think the request for the 'raw' data was coming from - show the crr/watts/elephant fart equivalents/whatever, just in raw, not the delta against a baseline that would appear to be somewhat contrived / synthesized/extrapolated.

In the first chart, "delta" is increasing as PSI is increasing (red line). I interpret this as "more watts are used" as the PSI increases.

In the second chart it's flat.

I always believed that until you hit the breakpoint, CRR would decrease with PSI then at breakpoint would go up. If neither of these charts are doing that, then we are close to, or beyond breakpoint ?

If we are at breakpoint at 65PSI (first chart), then maybe road surface is quite rough ?

Maybe I am reading this all wrong or I don't understand (wouldn't be a first)

Hang on, too many charts and too many red lines ;-) This was the chart I was referring to.

Yes, but look at first chart (and my post just above)

if you look at only that instance in zipp's recent testing on which you have reliable data - 25mm tires on an 858 NSW - the optimal pressure is low. this argues for exactly what you're saying: that the road surface on zipp's rolling road must be course. but it isn't course.

so i fall back on 2 possible explanations: first, that there's something else in the rolling road. vibration that occurs in the whole testing apparatus, causing the rider to unweight, even just a little, but constantly, making the rolling road test as if the surface is coarser than it is. but this is a wild guess on my part. second, that i would bet that almost all of the data that you, tom, al, have is based on tubed systems. please correct me if i'm wrong. zipp's is tubeless. is there a different optimal pressure for the same tire on the same road, tube v tubeless? is the inner bead width of the rim on the testing you've done similar to the [old] zipp 858 NSW, realizing that that wheel is not that old (i think from memory it was debuted at the 2019 IM).

so, you're right in the data you're looking at, if i understand you correctly. what i write above is what's running around in my head as i try to square that circle. but...

the point i'm trying to make, perhaps poorly, is what do yo do with the data you do see? that regardless of road surface, regardless of pressure, the new wheel with the 28mm tires outperforms the old. consistently. by 4 or 5 watts per pair. i take [what i think is] your point, that if this delta only exists on rough, chip & seal pavement, does it exist in kona? 4 watts is a big gap to close. how smooth does a road need to get in order for the old wheel to be as fast as the new? that's the wager.

---

Dan Empfield
aka Slowman

i don't know if this is helpful, but the red line as the baseline is the paradigm used for the wind tunnel test data that zipp offered.

that's not the paradigm for the rolling resistance tests. if you look at the red line on the Crr chart for the 858 that red line is, well, "unstraight." what i pointed out in my article, and what marcag also mentions if i understand him right, is that line should *not* be straight in the Crr chart for the 808. as the pressure goes up and down that line should change. but it doesn't.

---

Dan Empfield
aka Slowman

Agreed


I think we are converging :-)

If we saw the chart on a smooth road would there still be 4 watts at 110-110 PSI ? Or would it be 2 ?

Then if we had the aero penalty for the 28 tire would that be 2w there ?

there are too many things wrong with this data to have been generated for marketing purposes. as marcag and i have been discussing, the ideal pressures are too low during the 858 Crr test. the Crr doesn't change enough when pressures change during the 808 test. this is not the set of graphs you publish if what you want is to produce a marketing buzz.

i traveled back to indy, spent some time there, looked at their field testing, talked to their engineers, toured about every square foot of the factory, and maybe i'm fooled. maybe there's some 3D judo going on here and they intentionally through a curve ball by producing imperfect graphs in order to convince me that it wasn't marketing hype when in fact it was. but you have to be kind of a conspiracy theorist to think so.

why are there no error bars? because it's not a white paper. it's not intended for peer review. in which case, you might say, don't publish anything at all. then zipp would be criticized here, as manufacturers always are, when they launch a product without accompanying data.

by the way, off topic, but i've always wondered, perhaps everybody knows this but me, sorry if that's the case, but are you part of the hugi family? that hugi family?

---

Dan Empfield
aka Slowman

for sure, if 110psi was the best pressure we'd have a very different outcome. but in that case - on that surface - perhaps 25mm might be too wide.

this is one reason why i have contended - to the consternation of several of our partners who help keep the lights on - that hookless is a 28mm tire phenomenon. and up. 28mm the minimum tire width. once you go hookless you wed yourself to the idea that it's at least 28mm you'll be riding. otherwise, the pressures are too high. you can't ride hookless beyond 5 bar. so, what road is that? to circle back to your point. is hookless and 28mm kona? or paris roubaix?

in point of fact, paris roubaix is 30mm or 32mm. tubeless. that's what the pro peloton rides right now. off topic, but pidcock did that descent, unless i misremember, on a new dura ace wheelset, tubeless, inner bead width of 22.4mm. when i read what was written about that descent it was assumed that he rode a 25mm tire, but i think that's a big claim to make. if i had $5 i'd take that bet and wager he was on 28mm.

---

Dan Empfield
aka Slowman

  

On the aero side of things, I found what you wrote interesting

"However, as I read the notes on the test the one notable item is that 25mm tires were used throughout in the wind tunnel tests. (Not in rolling resistance, but wind tunnel.) I don’t have any wind data using a 28mm tire on either of these new wheels. While it would have made some sense to test this wheel with a 28mm tire, Zipp chose the 25mm size – in fact they used Zipp Tangente tires for their aero testing – because of all the historical data they have in the tunnel that used this tire."

I had a flash back to this



I wonder if we have hit the peak of aero on wheels

That's one way to look at it. But I'd argue Zipp would phrase it slightly differently as finally looking at wheels as a system optimization problem instead of just an aero optimization problem.

I was surprised with this forum's past wind tunnel test fetishization, this recent testing wasn't covered here (afaik). Which at least indicates the one maxim that if you don't want to agonize over which wheelset to buy, just buy some HEDs, as they've finished 2nd-3rd in about every wind tunnel test ever, as far as I can remember. The DT Swiss wheels were an interesting outlier, though.

Though I agree this set of data by Zipp is perplexing, and raises more questions than are answered.

Hoo boy...I go out for a ride on a Sunday morning and all confusion breaks out... ;-)

So...all this talk about "what do these charts even mean?" is ALL because of the data presentation. That is, the use of the "delta" charts with no clue as to what the reference curve looks like, and with poor explanation. Is the one curve flat because that's the delta to itself? OK, so what does the actual curve look like. THIS is why I said it was a "dog's breakfast" trying to make heads or tails of what is actually being presented.

In addition, being able to discern power values on their "rolling road" to within 1W...I'm highly skeptical. It's not that I don't' think they're good engineers, it's that I know how difficult it is to get that kind of precision with simple roller tests. Their "rolling road" is just going to ADD uncertainty, not to mention that the flat surface doesn't take advantage of the "amplification" effect of rollers in determining tire hysteresis properties. Show the setup and why we can have confidence in the results.

And then they say the aero data is with a 25mm tire (for comparison sake to past data) but then don't also show what it is with the 28? I'm sorry...that's just not kosher...especially if they say they don't show the Crr data on the older wheels with the wider tires. Sure, it was "optimized" around the 25mm, but THAT is an aero effect, not Crr. It reads like they're cherry picking the results.

I really don't know what to say. Zipp data used to be quite transparent (relatively speaking, for the industry), but this whole thing just doesn't make sense. Looks like I'm not the only one to feel that way, judging by this thread...

Here's an important thing to remember about the chart above from Silca...it shows BOTH roller data AND field test data...and, for similar surfaces, they MATCH up until the breakpoint pressure. That shows that roller testing is a GOOD proxy for on road Crr for most typical roads, and not only is a "quick and easy" way to determine this, but also accurate. Here's the thing...if Zipp's rolling road data isn't matching either roller testing trends OR field test data (as you mentioned, they said they weren't happy with their results and haven't shown them), then maybe that says something more about the applicability of their treadmill? It IS the outlier apparently...

---

http://bikeblather.blogspot.com/

Hmmm...maybe I'm missing something...the y-axis is labeled as Crr, and the values are of a range I'd expect for the data. Please educate me on what's incorrect. Genuinely curious.

---

http://bikeblather.blogspot.com/

The answer to that is demonstrably "no" (when using a latex tube). Asked and answered, your honor. Plenty of data to confirm this.


The point I'm trying to make, perhaps poorly, is based on the data presented, I can't make ANYTHING with the data I see. And that's because, IMO, the data have been poorly presented.

How about THIS wager, how much slower aerodynamically does that new wheel get with the 28 on there, instead of the 25 in the data presented? Is it 4-5W? More? This is a big reason I endeavored to present a combined aero+Crr estimate when I showed the results from the "Win Tunnel Playtime" on my blog.

---

http://bikeblather.blogspot.com/

Exactly...if you are coming up with data that's supposedly "different" than what has been seen previously, don't change the format of the data presentation. Especially if you don't show the baseline curve that you're normalizing to on a "delta" plot...it all just looks like something is trying to be hidden IMO.

---

http://bikeblather.blogspot.com/

I had to look twice, but notice the y axis. Notice .002, .004, .006 then .007, .009....