Biorow - 1x Boat testing in indoor tank

[thomas....@googlemail.com]

Biorow have released their findings from running tests on 4 different singles manufacturers (Empacher, Wintech, Filippi and a local manufacturer Nowing) where they measured the boats when they were being sculled down the tank over 200m (4 runs at 22, 26, 32 and 42) and when they were being towed

http://www.biorow.com/RBN_en_2015_files/2015RowBiomNews04.pdf

Their findings are interesting as it shows that both for this sculler and when being towed the Wintech shows the lowest drag and that at the lower rates the Wintech and Empacher do worse than the Fillipi and Nowing make.

I also think its interesting to see that the Empacher does the best with regards to blade efficiency, which means despite beign ranked 3rd in average power it had the highest propulsive power which resulted in the highest average boat speed, level with the Wintech. Assuming the same blades and rig was used, perhaps this athlete was the most "comfortable" in the Empacher boat which meant he could get the most out of it?

IMO it would be good to see this testing being done more times with other scullers to see if this is due to the hull shapes or is just due to this particular scullers style/technique

[Kit Davies]

Sorry, OT, but I'm most impressed by the idea of having a 600m indoor
rowing lake. I know clubs who row on less water than that. Think of the
possibilities!

As for the results: so in ideal controlled conditions, one boat will
produce an advantage of 1s over 2K. My conclusion is what most people
know already, that it's the horse not the chariot.

Kit

[thomas....@googlemail.com]

One further point, I note that the local manufacturer NoWing did the best with respect to boat speed vs propulsive power, so although it was slower than the others in the sculling tests it also had the lowest propulsive power of all the boats. So if the sculler were able to produce the same power in each boat an, the NoWing would be the fastest so that would be worth investigating why the power was so far down in that boat (uncomfortable? Sculler bias?)

[sully]

On Monday, May 11, 2015 at 8:43:15 AM UTC-6, Kit Davies wrote:
> On 11/05/2015 14:40, thomas....@googlemail.com wrote:
> > Biorow have released their findings from running tests on 4 different singles manufacturers (Empacher, Wintech, Filippi and a local manufacturer Nowing) where they measured the boats when they were being sculled down the tank over 200m (4 runs at 22, 26, 32 and 42) and when they were being towed
> >
> > http://www.biorow.com/RBN_en_2015_files/2015RowBiomNews04.pdf
> >
> > Their findings are interesting as it shows that both for this sculler and when being towed the Wintech shows the lowest drag and that at the lower rates the Wintech and Empacher do worse than the Fillipi and Nowing make.
> >
> > I also think its interesting to see that the Empacher does the best with regards to blade efficiency, which means despite beign ranked 3rd in average power it had the highest propulsive power which resulted in the highest average boat speed, level with the Wintech. Assuming the same blades and rig was used, perhaps this athlete was the most "comfortable" in the Empacher boat which meant he could get the most out of it?
> >
> > IMO it would be good to see this testing being done more times with other scullers to see if this is due to the hull shapes or is just due to this particular scullers style/technique
> >
> Sorry, OT, but I'm most impressed by the idea of having a 600m indoor
> rowing lake. I know clubs who row on less water than that. Think of the
> possibilities!
>

Indeed!

I've been daydreaming about this for my cold-water coaching pals for a couple decades.

What would be even better would be to turn it into a flume, nothing really fast just that the upstream pace would take a lot longer, but that you could control the boat when you turn around on the downstream leg. Coach could coach by walking alongside!

I thought that would be a good use of the Stanford Linear Accelerator when it sunsetted it's High Energy Physics experiments, but the doggone physicists figured out how to use it for photon science...

[Leo]

On Monday, 11 May 2015 23:10:50 UTC+9:30, thomas....@googlemail.com wrote:
> Biorow have released their findings from running tests on 4 different singles manufacturers (Empacher, Wintech, Filippi and a local manufacturer Nowing) where they measured the boats when they were being sculled down the tank over 200m (4 runs at 22, 26, 32 and 42) and when they were being towed
>
> http://www.biorow.com/RBN_en_2015_files/2015RowBiomNews04.pdf
>

IMO, this is Valery's best ever report. He is to be congratulated on his
scientific approach, and he also deserves a big bear-hug for making the
results available to the rowing community.
Simply magnificent!

[Eberhard Nabel]

Caught just a glimpse of the paper. The smaller shells perform better than the bigger ones. Hard to derive any conclusion from that, which make is better.

[wmar...@gmail.com]

It is interesting to see the results and initially it looks good for WinTech, but as Eberhard remarks, the boats are for different weight scullers. It may be that the study was done with what was available rather than boats built for similar crews - the latter would be preferred. Wonder if the manufacturers will supply equipment... Hmm. An expensive study - Carl Douglas, Hudson, Filippi, Empacher, WinTech, Vespoli, VanDusen, Sykes, KIRS, and many more, would make a complete study?

[carl]

For now the most important thing is that this study has been done. It
is a welcome start. The manner in which it has been done is nothing
special - it's as it should have been done - but as ever in science we
will gain confidence only after these test results have been repeated &
validated (if that's the case) by others.

The immediate value of these tests is that they may open eyes just a
little among that majority sheeplike tendency which believes that the
equipment most heavily promoted & most widely favoured by national teams
must, automatically, be "the best" - which is rather like believing in
fairies.

I have always advocated, on RSR & more widely, that equipment should be
objectively evaluated before purchase. Meanwhile huge sums continue to
be spent in acts of blind faith following what others presume, without a
shred of objective supporting evidence, must actually be best.

Cheers -
Carl
--
Carl Douglas Racing Shells -
Fine Small-Boats/AeRoWing Low-drag Riggers/Advanced Accessories
Write: Harris Boatyard, Laleham Reach, Chertsey KT16 8RP, UK
Find: tinyurl.com/2tqujf
Email: ca...@carldouglasrowing.com Tel: +44(0)1932-570946 Fax: -563682
URLs: carldouglasrowing.com & now on Facebook @ CarlDouglasRacingShells

[Leo]

On Tuesday, 12 May 2015 23:07:11 UTC+9:30, carl wrote:
> On 12/05/2015 13:16, wmar...@gmail.com wrote:
> > On Monday, 11 May 2015 09:40:50 UTC-4, thomas....@googlemail.com wrote:
> >> Biorow have released their findings from running tests on 4 different singles manufacturers (Empacher, Wintech, Filippi and a local manufacturer Nowing) where they measured the boats when they were being sculled down the tank over 200m (4 runs at 22, 26, 32 and 42) and when they were being towed
> >>
> >> http://www.biorow.com/RBN_en_2015_files/2015RowBiomNews04.pdf
> >>
> >> Their findings are interesting as it shows that both for this sculler and when being towed the Wintech shows the lowest drag and that at the lower rates the Wintech and Empacher do worse than the Fillipi and Nowing make.
> >>
> >> I also think its interesting to see that the Empacher does the best with regards to blade efficiency, which means despite beign ranked 3rd in average power it had the highest propulsive power which resulted in the highest average boat speed, level with the Wintech. Assuming the same blades and rig was used, perhaps this athlete was the most "comfortable" in the Empacher boat which meant he could get the most out of it?
> >>
> >> IMO it would be good to see this testing being done more times with other scullers to see if this is due to the hull shapes or is just due to this particular scullers style/technique
> >
> > It is interesting to see the results and initially it looks good for WinTech, but as Eberhard remarks, the boats are for different weight scullers. It may be that the study was done with what was available rather than boats built for similar crews - the latter would be preferred. Wonder if the manufacturers will supply equipment... Hmm. An expensive study - Carl Douglas, Hudson, Filippi, Empacher, WinTech, Vespoli, VanDusen, Sykes, KIRS, and many more, would make a complete study?
> >
>
> For now the most important thing is that this study has been done. It
> is a welcome start. The manner in which it has been done is nothing
> special - it's as it should have been done - but as ever in science we
> will gain confidence only after these test results have been repeated &
> validated (if that's the case) by others.
>
> The immediate value of these tests is that they may open eyes just a
> little among that majority sheeplike tendency which believes that the
> equipment most heavily promoted & most widely favoured by national teams
> must, automatically, be "the best" - which is rather like believing in
> fairies.
>
> I have always advocated, on RSR & more widely, that equipment should be
> objectively evaluated before purchase. Meanwhile huge sums continue to
> be spent in acts of blind faith following what others presume, without a
> shred of objective supporting evidence, must actually be best.
>

Well put, Carl.
I agree that the method of tank testing is quite standard and that
it should be done more often.

Of course, repeating the tests would be a bonus, but that would be
very expensive and not without its own difficulties. The ITTC tried
to conduct tests with a destroyer model hull (actually 2 geosims) at
more than 30 tanks around the world. It took several years and there
were all sorts of problems unravelling the (double blind) results
and moving the hulls around the world.

Some colleagues also pointed out that it would have been better if
the tests were done at a couple more speeds (especially between 4 m/s
and 5 m/s) but I suspect that would have cost a lot more.

I also have some disagreements with Valery's way of calculating a
"gross rowing drag factor" because it assumes the drag is the same
at the same speed, when in fact it is different depending on the
position and poise of the sculler in the shell. That's a minor
point for me: the experiments are what are important.

So while I do have some criticisms, I'm just delighted that Valery
has made the results public. There are only a few others who have
done that (e.g Wellicome and Scragg and Nelson). He also deserves
praise for using full-scale models and a large tank where bank
reflections and depth effects are inconsequential.

Sadly, I suspect that there will be those who discount the effort
and "cite" their own anecdata. Like the poor, unscientific fuckwits
will always be with us.

[Stewie]

Leo

I also welcome this scientific approach, especially the towing tests, but regarding their validity I think you touched on something important here:

> I also have some disagreements with Valery's way of calculating a
> "gross rowing drag factor" because it assumes the drag is the same
> at the same speed, when in fact it is different depending on the
> position and poise of the sculler in the shell. That's a minor
> point for me: the experiments are what are important.

It is not just the transfer of weight that affects the drag, but also the oscillating variation in hull speed through the stroke cycle. Drag on the shell needs to be taken as an average over a cycle (or perhaps thought of as a net impulse per cycle rather than a simple force or drag coefficient) to represent what happens in a real rowing stroke. That is much more difficult to accurately replicate with towing equipment.

Carl has discussed the importance of hull speed oscillation drag here and in articles elsewhere. Its effective elimination may explain discrepancies between towed values and rowed values for drag.

Cheers
Stewie

[Leo]

On Wednesday, 13 May 2015 19:19:50 UTC+9:30, Stewie wrote:

snip

>
> I also welcome this scientific approach, especially the towing tests, but regarding their validity I think you touched on something important here:
>
> > I also have some disagreements with Valery's way of calculating a
> > "gross rowing drag factor" because it assumes the drag is the same
> > at the same speed, when in fact it is different depending on the
> > position and poise of the sculler in the shell. That's a minor
> > point for me: the experiments are what are important.
>
> It is not just the transfer of weight that affects the drag, but also the oscillating variation in hull speed through the stroke cycle. Drag on the shell needs to be taken as an average over a cycle (or perhaps thought of as a net impulse per cycle rather than a simple force or drag coefficient) to represent what happens in a real rowing stroke. That is much more difficult to accurately replicate with towing equipment.
>
> Carl has discussed the importance of hull speed oscillation drag here and in articles elsewhere. Its effective elimination may explain discrepancies between towed values and rowed values for drag.
>
> Cheers
> Stewie

Of course that variation is very important, Stewie.
What I meant is best illustrated in Figure 39 on page 60 of the FIRM
manual.
http://www.boatdesign.net/forums/attachments/design-software/99036d1427770325-free-internet-rowing-model-2-32-released-firm_manual.pdf

Not only does speed vary during the stroke, but the drag for the same
speed is different. For example, the speed halfway through the drive
might be exactly the same as halfway through the recovery, but the
drag at that speed can be different. That is due to unsteady effects,
(mostly surge and pitch), but also because the air drag is different.

To Valery's credit he did test the hulls with real scullers propelling
the boats, and also by towing the hulls at constant speed.
My (very minor) criticism is the use of a constant drag factor in the
estimation of propulsive power for the former case.

And I agree it is tough to replicate! Day et al described their
experiments in a tank and in the field in the paper:
"Realsitic evalaution of hull performance for rowing shells, canoes,
and kayaks in unsteady flow",
J. Sports Sciences.29:10, 2011, pp 1059-1069.

I'd love to see more experiments like that, but it is an enormous
task.

[thomas....@googlemail.com]

> To Valery's credit he did test the hulls with real scullers propelling
> the boats, and also by towing the hulls at constant speed.
> My (very minor) criticism is the use of a constant drag factor in the
> estimation of propulsive power for the former case.

Agree, although the towing element of the study is interesting in so much that it go against the trend of thinking regarding boat makes, personally I find the data relating to the tests involving the sculler to be more interesting as it more accurately shows the boats "in use" and if you hunt for data where the rate and Propulsive Power were similar between two different manufactures (i.e rate 26 between Wintech and Empacher or rate 22 between Filippi and NoWing) the boat speeds are almost exactly the same.

[carl]

What we're looking for is, I suppose, a way of correlating towing test
results with rowing test results.

As you say, Stewie, speed & pitch oscillations seriously complicate any
analysis: you have the boat interacting cyclically with the boundary
layer (which has considerable mass) & the resulting unsteady BL
interacting with the rest of the body of water. A bit of a challenge.
And when you've seen how very different can be the boat speed (& pitch)
oscillations from sculler to sculler, the correlation challenge only
grows - each sculler will lead to a somewhat different correlation.

A few years study in all that!

[Leo]

I have tried to account for most of those factors in my model, but
it requires a lot of input. Users must supply limb lengths, body angle
regimes, oar lengths, forces and angles, hull geometry, seating
positions etc etc. That can be very tedious, especially for eights.
It's a bonus when experts measure and collate most of the data for
you!

Fortunately, I had good on-water data for elite rowers. It would be
tougher to get good correlations with less capable rowers who are far
less consistent. Even top class rowers are not all that consistent at
the start of a campaign after a lay-off, but when they are nearer to
competition they are sometimes preternaturally consistent, e.g. 20 or
more consecutive strokes at 36 spm +/- 0.2 at most. I don't think I
can walk that steadily!

I'm still thinking of how to include the effect of roll on total drag
(including the boundary layer). Because rowing shells have cross-sections
that are quite simple and very smooth, I'm tending to think might not
really be all that important.

[tcyr...@gmail.com]

"Damn the photons, full speed ahead!"

On Monday, May 11, 2015 at 1:06:43 PM UTC-5, sully wrote:
> On Monday, May 11, 2015 at 8:43:15 AM UTC-6, Kit Davies wrote:

[Steve S]

As I recall, Ted Van Dusen was able to get at least one full-sized single towed at the David Taylor Model Basin. It would be wonderful if he was inspired by the Kleshnev example to release his data as well.

[Leo]

On Thursday, 14 May 2015 11:21:27 UTC+9:30, Steve S wrote:
> As I recall, Ted Van Dusen was able to get at least one full-sized single towed at the David Taylor Model Basin. It would be wonderful if he was inspired by the Kleshnev example to release his data as well.

It would be good, but because it is an expensive exercise there are good commercial reasons not to release it.

I think that on-water testing with a variety of rowers is better than tank tests at a steady speed. Unfortunately, it's not always possible to get very calm conditions to get good comparisons and correlations.

[John Greenly]

On Thursday, May 14, 2015 at 12:44:51 AM UTC-4, Leo wrote:

> I think that on-water testing with a variety of rowers is better than tank tests at a steady speed. Unfortunately, it's not always possible to get very calm conditions to get good comparisons and correlations.

Leo,
I wonder also, how well such results on nice flat water would predict how different hulls would do in rougher conditions. Some boats seem to have a reputation for being faster when it's flat, or rough. Hull stability as well as drag of course can enter significantly into that comparison.

--John

[Leo]

Yes, it is an important issue.

I have very good code for predicting (ambient) wave effects on ships, and it works well for thin slender hulls like rowing shells.

The difficulty is in getting good, reliable estimates of the wave spectra. These are available for many locations that are appropriate for ships, and some that are reasonable for larger yachts, but there is very little that applies to rowing shells. Small waves that can impact on the performance of small boats like rowing shells and kayaks are immaterial for ships. There is little incentive to collect wave statistics that are relevant to small boat performance. (That might have changed in the last few years, but I haven't had the time to look into it.)

IMO top class teams should have two boats: one for calm conditions and one for rougher weather, but that is simply beyond the budgets of most countries.
Maybe the answer is for elite crews to be trained to handle all conditions and not make excuses like "we lost because the weather didn't suit us". :)

[wmar...@gmail.com]

On Thursday, 14 May 2015 13:15:11 UTC-4, Leo wrote:
(snip)

>
> IMO top class teams should have two boats: one for calm conditions and one for rougher weather, but that is simply beyond the budgets of most countries.
> Maybe the answer is for elite crews to be trained to handle all conditions and not make excuses like "we lost because the weather didn't suit us". :)

What?!?!? Common Sense?!?!? Stop! Not allowed!
;-)

[s...@ku.edu]

Leo,

How would comparing the MW Wintech (length= 7.9m) to the HW Emp 10 (8.2m) and Filippi 14 (8.33m) influence the drag results?

Steven M-M

[Leo]

I don't know because I don't have the offsets of all the hulls, or
the weight, limb lengths, force-time profiles etc of the scullers
Valery used in his tests, or the way they move their weight during
the stroke. All of those factors (and many others) affect the drag
at different instants during the stroke. In short, some hulls will
suit some scullers better than others.

I have included 12 1x hulls with FIRM. Two of these are very
similar to the Empacher and Filippi hulls used by Valery in his
tests. I measured the hulls myself, but I didn't want to identify
them by manufacturer's name.

It's easy to change the hull in the 1x examples included in FIRM
by changing the hull in the input files. Run the program with
different hulls and then compare the speed, drag components, power
components etc.

It's also easy to compare the effect of different hull lengths or
widths etc by changing values in the input files. Of course, that
won't necessarily correspond to an existing hull, but it will give
you an idea of the effect of using a longer hull while keeping the
total weight constant and other factors constant. There are many
other simple comparisons that can be done this way, e.g. effect of
air and water temperature, wind speed and direction, force profiles
etc. But you have to invest time in learning the program's nuances.

You can see some of the output in the 1x examples for lightweight
and heavyweight men and women in:
http://www.boatdesign.net/forums/attachments/design-software/99037d1427770364-free-internet-rowing-model-2-32-released-firm_examples_singles.pdf

The examples also show on-water measurements of speed and
acceleration for the example scullers at some nominal stroke rate.
I have a lot more data for these and other scullers, but the
examples and documentation already run to over 140 pages, so I
thought it best to not include more yet.

If you are really keen, then the same hull offsets and principal
dimensions of hulls included with FIRM can also be used in other
free codes like "Flotilla" to estimate hydrodynamic drag and squat
at steady speeds. Flotilla and FIRM are available for free from
my web page: http://www.cyberiad.net/leo.htm

[s...@ku.edu]

Thanks, Leo. My mistake in asking for a univariate answer to a multivariate problem. I find your summary "Power Flow Charts" well worth puzzling over, although I'm not confident I grasp the details. Just to double check my understanding (and risk exposing my misunderstanding): the rower's work input, A, is divided into ~ 80% work on handles + ~20% system momentum. Am I correct that this second energy input is primarily the use of the sculler's body movement during the drive and recovery? I further assume (but please correct me) that most of this kinetic energy derives from the mass of the rower moving during the recovery.

I was also struck by the differences between the 4 scullers on the "hull velocity and crew CG velocity" sub figures. This is one area they may define the individual sculler's technique. Can you derive any advice from this analysis re efficiency?

Steven M-M

[Leo]

Steve:
1. Sorry to have given such a prolix reply to your previous question,
but there are no simple answers when it comes to the total drag of a
oar/rower/hull system as there are for hulls on their own in steady
flow. Also, as somebody else pointed out, the "design" weights of the
hulls that Kleshnev examined are different, which makes comparisons
difficult if not meaningless.

You have to remember that some hulls will also perform differently
at low stroke rates (around 15-20 spm) and at racing rates (> 30 spm).

2. Yes, the majority of the 2nd component is due to the oscillation of
the system. The values I calculate for the internal losses (Box G)
are significantly lower than Atkinson's estimates, but they are very
similar to those that Kleshnev and others find in their work.
Unfortunately, Atkinson does not give explicit formulas for his
calculation, a point made by Hofmijster and others in their papers.
My values are given by equations (54) and (55) in the paper.

3. I'm not sure what you mean by the 4 scullers and the "hull velocity
and crew CG velocity" sub figures. Can you give me the number of the
figures on the manual?

[s...@ku.edu]

The right panels of Figures 1, 11, 20, & 29. The differences between the W1x and LW1x caught my eye. My visual interpretation is that for the W1x hull velocity continues to increase through the recovery up the point the sculler's weight shifts to the foot stretcher. For the LW1x the hull velocity remains relatively constant over the recovery until just before the catch. The M1x and LM1x patterns were distinct, but more similar. I am focusing on this in part because of Kelshnev's Jan 2015 analysis the variation of boat velocity and its effect on energy and speed over a 2k. http://www.biorow.com/RBN_en_2015_files/2015RowBiomNews01.pdf

Steven M-M

[Leo]

Steven:
You are correct that the velocity curve for the LW1x is unusual,
as I mention near the bottom of page 8 of the singles examples.
(I have other on-water velocity data for the LW1x class which is
not as constant during the recovery.) That said, it is interesting
that it is possible for scullers to produce an almost flat velocity
curve for a large part of the recovery.

I also agree that it might "define" in part a sculler's technique
but, as I mention in the main manual, I do not think that FIRM
should be used to optimize technique. There are far too many
subtleties constituting "technique" which are beyond my model.
Those subtleties are better addressed by experienced coaches.
My main interest with FIRM was to capture as well as I could the
hull velocity and acceleration given the forces, oar angles and
other on-water data.

[s...@ku.edu]

Thanks, Leo. The difference in SR for the LW1x (~28) versus the other three (~35) likely explains the difference. I suspect that 'Laura's' velocity curve would be similar to the others at SR 35. Again just eyeballing the figures, the LW1x had the smallest difference, 1.7, between max speed (4.5) and min speed (2.80) compared to the other scullers. This lead me to wonder if the flatter recovery velocity curve might be desirable for longer head races which are typically rowed at similar rates.

SMM

[carl]

And in that you have done a remarkable job, Leo. I, and clearly some
others, are very much indebted to you. Thank you.

May I now revert to the matter of shell testing?

Yes, that tank testing is inevitably a costly exercise - it's a vast
piece of kit to take over at who knows how many bucks/day. And, yes,
you require those controlled conditions for constant-velocity loaded
hull measurements. However, for the live sculling tests you add in all
manner of variables - rig, actual hull size sculler preferences &
prejudices, etc.

Next, since we're discussing racing craft, I am unsure of the direct
relevance of low-rate, low speed evaluations. Who cares how a shell
performs at low power when everyone is busting a gut to win at high
power and rate? Yes, it may be interesting, but to me it is perhaps not
the best use of costly testing time. Of greater interest might be the
relationship between rating & speed around race pace, & I say that
advisedly since we should perhaps be looking at Strouhal-related
frequency effects resulting from oscillatory pitch & surge inputs.

But, stepping back a little further into what one might call the real
world with realistic costs, what in reality is so wrong with testing a
bunch of shells with a bunch of scullers on still waters of constant
depth & in still conditions as, ultimately, we care only about which
shell performs best under race conditions? Yet these are the very tests
which rowers seem so reluctant to conduct before purchase, apparently
fearing to reach the "wrong" (in the eyes of the herd) conclusion.

You'd systematically switch boats & scullers &, ideally, would
camouflage the boats to reduce rower-preconceptions - since this
exercise could be seen as somewhat akin to seat racing, & we all know
that rowers will fix seat-race results to end up rowing with their
mates. You'd also want shells in sizes rated by their makers as being
above, at & below their trial horses' weights (where these options
actually exist) to resolve manufacturer misjudgements on size. And no
sculler should have even the least inkling of their recorded times.

I'm not saying this would be easy, but it is feasible. And I'd venture
to suggest that the results, while lacking the analytical intricacy of
the tests we've been discussing hitherto, might well be more meaningful.
And be achieved at far less cost.

Please remember that you heard this from a racing shell builder who
likes his clients to place measured performance high among their
priorities for shell selection.

[Leo]

Thanks, Carl.
I hope some people find FIRM to be a useful research tool, one that can
give reasonable estimates of the effect of changing one variable at a
time. I hope they also take note of my warnings in the manual about what
the program can do, and what it should not be used for, e.g. "optimising"
technique.

(I know you know this, but I'm warning others who aren't as savvy...)
Kleshnev's intentions in his tests should be kept in mind. He set out
to validate his own on-water measurement system and modelling. He did
NOT set out to show the superiority of one hull over the others and I
hope nobody is silly enough to think that his tests "prove" anything
of the sort.

Your comments about on-water testing are astute, and I know some
people who are planning to try some of your suggestions. Sadly, I don't
think they will go as far as trying to anonymise the boats to minimise
biases. (I didn't realise that rowers sometimes "rig" seat races to
end up with their mates, but it's not surprising.)

I also agree completely about tests at low stroke rates. I have
enormous amounts of on-water data for stroke rates between 16 spm and
28 spm. I don't consider anything below about 28 spm, with the
exception of the LW1x (at 27.9 spm) I included as an example because
it is so interesting. Coaches might find low stroke rate measurements
useful for their own reasons, but for my purposes they aren't relevant
to real races. I am surprised at how much effort is expended in
measuring performance at those low stroke rates. I suspect many coaches
give them a cursory glance and never look at them again.

There's no doubt that on-water testing is difficult, but many teams
conduct them often and routinely to select crews and to assess the
effects of a variety of methods and techniques. The difficulties
you mention about comparing boats are not insurmountable, so I'm at
a loss to understand why boats are not assessed as part of the total
strategy for world championships and the olympics. Maybe it's easier to
believe advertising guff, or opinions? Maybe it's easier to use the same
boats as other teams to avoid being blamed for a loss on a poor boat
choice?

[John Greenly]

Leo,

I'd like to add my thanks for your modeling work. I am going to try out FIRM and I'm certain it will be extremely valuable to help unravel some things I've been thinking about.

Since we are on the subject of testing hulls to measure drag, I'd like to bring up, and hope to have your reaction to, something I suggested here a long time ago that I think i did a bad job explaining. You've obviously thought through all the relevant issues in developing your model.

My thought is this: if you look at the hull as a single, rigid object, it has mass, and there are forces on it. The forces are related to the mass and acceleration by Newton's law. There are only three significant forces acting axially (fore-and-aft) on the hull causing its acceleration (assuming the seat slide is good). These forces are: the axial component of the force exerted on the gates by the oars, the axial component of the force exerted by the feet on the stretcher, and the total drag force. The sum of these forces equals the hull mass times its acceleration. The instrumentation is now available to measure the gate and stretcher forces and the acceleration, so we should be able to directly solve for the drag force, as a function of time through the stroke, during actual rowing. Is somebody already doing this? I think Valery Kleshnev collects all the necessary data.

--John

[s...@ku.edu]

2 thoughts:

1) Like many masters rowers, I consider 4-5k head races the focus of my competition. So knowing how a 1x responds and what makes effective technique at SR 28-30 is important to me and many others.

2) While the analysis presented in April '15 RBN is best seen as a test of the protocols, Kleshnev does make specific between boat speed comparisons: "... above differences in the drag factor would give the advantage to WinTech boat 1.02s over Empacher, 3.06s over Nowing and 3.84s over tested Filippi." You can add all the technical caveats you like, but the plain English interpretation is that one boat was faster. If I was doing peer review, I would tell the editor "revise and resubmit."

SMM

[Leo]

On Monday, 18 May 2015 04:29:47 UTC+9:30, s...@ku.edu wrote:

snip

> 1) Like many masters rowers, I consider 4-5k head races the focus of my competition. So knowing how a 1x responds and what makes effective technique at SR 28-30 is important to me and many others.
>
> 2) While the analysis presented in April '15 RBN is best seen as a test of the protocols, Kleshnev does make specific between boat speed comparisons: "... above differences in the drag factor would give the advantage to WinTech boat 1.02s over Empacher, 3.06s over Nowing and 3.84s over tested Filippi." You can add all the technical caveats you like, but the plain English interpretation is that one boat was faster. If I was doing peer review, I would tell the editor "revise and resubmit."
>

Steven:
1. In relation to the LW1x results, I should have added that Lara's
internal losses (Box G) are only 4% of the total, which is the smallest
fraction I have found so far. The larger boat to rower mass also
contributes to that low fraction.
As you point out, it would probably be larger at higher stroke rates.

Thanks also for reminding me that lower stroke rates are applicable to
longer events. I have been so focussed on 2k events that I forgot about
other types of racing. (I pulled all my work on fixed-seat rowing
because I didn't have good data.)

2. Point taken about Kleshnev's comparisons. I'm still wondering if
some of the averages he tabulates are meaningful.

[Leo]

Thanks John.
Yes, it is possible to solve for the total drag using, for example, the
equations in section 6 of the main manual. Others e.g. Hofmijster et al
do try to solve for the drag. I prefer to split the drag into components
so I can see the various proportions.
Hofmijster's thesis is available at:
http://www.hemus.nl/maand/2010/mechanics_rowing.pdf

I don't like the use of the speed-squared relation. Reynolds number
effects mean that an exponent other than 2.0 is often better. As I
mentioned earlier, using a constant area in the drag factor (or to derive
a drag coefficient) is also problematic.

[John Greenly]

On Sunday, May 17, 2015 at 11:22:58 PM UTC-4, Leo wrote:
> Yes, it is possible to solve for the total drag using, for example, the
> equations in section 6 of the main manual. Others e.g. Hofmijster et al
> do try to solve for the drag. I prefer to split the drag into components
> so I can see the various proportions.
> Hofmijster's thesis is available at:
> http://www.hemus.nl/maand/2010/mechanics_rowing.pdf
>
> I don't like the use of the speed-squared relation. Reynolds number
> effects mean that an exponent other than 2.0 is often better. As I
> mentioned earlier, using a constant area in the drag factor (or to derive
> a drag coefficient) is also problematic.

Thanks very much for the reference to that thesis, it deals with exactly some things I was puzzling about with respect to blade dynamics.

Yes, the 2.0 approximation has always been just that... there are many possible complications. One of the many things that would be interesting from deriving the drag as a function of time through the stroke would be to see how the drag at a given speed differs depending on whether the boat is accelerating or decelerating at the time. Of course, the hull pitch angle will also be different- so hard to isolate variables.

John

[mi...@mikebarker.co.za]

On Monday, May 18, 2015 at 8:10:16 PM UTC+2, John Greenly wrote:
> ... Of course, the hull pitch angle will also be different- so hard to isolate variables.

John,

If you are going to start using FIRM perhaps we can form a "users group" - always easier to learn together ? Everyone welcome !

You may find that an old smartphone with a 9-DoF sensor suite and an Android datalogging App like "SenseView" will measure hull pitch angle and more, albeit the raw un-filtered data.

[Leo]

FIRM includes hull surge effects on the boundary layer and thereby
on the viscous drag. Surge also affects wave resistance. Pitch and
yaw effects are also included on the wave resistance.

I have on-water measurements of yaw, pitch and roll for the examples
in FIRM but, as I mentioned earlier, I decided to stop including
more when I got to about 140 pages of documentation.

I'll return to FIRM in a few months. In truth, I've had enough of
rowing modelling for a while and I'm happy to be working on nice
"clean" applied maths problems e.g. quadrature where there are
exact answers guided by theorems. :)