Sailonline.org - Online Sailing

Facebook

Login

Support Sailonline

If you haven't already - join the SAILONLINE YACHT CLUB!

Please also consider making a donation - all amounts are greatly appreciated!

Tweets by @sailonline

Board » Technical Discussion » Performance loss

Page: First Previous 2 3 4 5 6 7 8 Next Last

Posted by javakeda Posts 71 Registered at 2013-03-01 03:18:22	2013-10-14 19:06:44 - Re: Performance loss
	A2R: The maxBtSpd for 17kts is about 13kt, so not really close to the 16+kts needed based on the BtSpd rule alone. I think the Treff-Team site, noted in my prior post, really tells us a lot. I would love to know the complete history of that site. I suspect that it was developed to support both virtual and real life racing. VR is a likely virtual venue. Note the heading "Ocean Virtuel" above the 3D-graphics that show color-coded sail preferences. I really have to admire the work that went into this site. So, why the switch to sailonline? I don't want to speculate. I am just glad it happened. Meanwhile, here is an alternative hypothesis for the PERF change: The BtSpd rule applies, but there is an additional penalty for changes in COG that exceed X-degrees -- and we have to figure out what X is. Maybe something like: PERF=(100-BtSpd/2)*deltaDEGREES/XDEGREES If it looked like that, I'd bet on XDEGREES=90 (pi/2 in radians) coupled with a test on the sign of the TWA. If the sign changes, the deltaDEGREES is assumed to be at least 90. But all that is just a priori guesswork. I should be able to test both hypotheses in the same data gathering series. Back to the stopwatch.
Posted by A2 Posts 347 Registered at 2009-10-26 00:39:42	2013-10-14 19:35:26 - Re: Performance loss
	Example cited is consistent with the formula (I should have done the math before commenting) --- Last Edited by A2R at 2013-10-15 11:24:27 --- None so blind
Posted by kroppyer Posts 244 Registered at 2010-02-05 15:25:55	2013-10-14 20:34:58 - Re: Performance loss
	Ok, so this is the image, and raw data from tonight's practice race. I think we can conclude that performance recovery is linear, or at least close to linear (without steps). If anyone wants to have the PHP script I used, ask me. You'll need to know how to run PHP scripts, and how to install pear's System Daemon-1.0.0 Attachments perf-model-2.png (31.5 KB) perflog.csv (16.8 KB)
Posted by Rod Posts 350 Registered at 2010-06-24 22:30:48	2013-10-14 21:07:49 - Re: Performance loss
	May I suggest, Kroppyer, in all due humility etc etc ...We are attempting here to clarify and to simplify the perf loss and recovery phenomenon for the benefit of those among us (such as myself) who either are many years since their last mathematics course or have only a small knowledge of math. If processes or notations, that are not simplified or clarified, are introduced in this discussion, it just further adds mystification to a procedure that is already somewhat obscure to many SOLers. p.s. An example was the introduction of 'radians' instead of 'degrees' into the discussion. I know radians etc, but their extra convenience in calculations is not something I understand immediately. I can visualize the Captain, in an emergency, giving the order, "0.356 Radians of sinister wheel on!" and being answered by "HUH?" i.e. it is not maritime terminology...... --- Last Edited by Rod at 2013-10-14 21:15:34 --- If it breaks, it's not strong enough--if it doesn't, it's too heavy.
Posted by javakeda Posts 71 Registered at 2013-03-01 03:18:22	2013-10-14 21:59:58 - Re: Performance loss
	And may I suggest, Rod, with all due humility that we watch and learn from Kroppyer. Note that his data logging program is far more accurate than hand logging from the SOL client. It establishes, beyond a doubt in my mind, that the PERF recovery rate is linear and constant. Well, done, indeed! As for the math terms in use, nothing mentioned in this thread so far goes beyond what I learned in 10th grade -- in 1954. When I became interested in routing, I realized I needed a refresher course or two. I took a number of courses at Khan Academy. Here is the url: https://www.khanacademy.org/ The courses are free, and bite sized. I recommend this site to anyone who wants to polish their basic math skills. As for using the term 'radians'...I offer no apologies. If you want to use a spreadsheet to calculate VMG and such, you need to be able to use the RADIANS function to translate from degrees to radians. This is not an obscure term.
Posted by kroppyer Posts 244 Registered at 2010-02-05 15:25:55	2013-10-14 23:30:09 - Re: Performance loss
	Rod, thanks for the feedback on my way of communicating in this thread. I am trying to find the formulas behind the performance loss and recovery model. When I found a model that is close enough to the one SOL uses, I will describe it in a way that most people will understand it, even if they're not too good with maths, I want to put this performance model in the calculator of spinnacer. While I'm trying to discover these formulas I'm not really focussing on making it understandable to all. I'm merely sharing my findings and thoughts for those interested. I hope everyone will do the same. All contributions of other people are welcome, the more data/thoughts/discoveries, the better. I will try describe and explain a little more, but I'm still trying to find the actual formulas and then explain them, instead of explaining how I'm trying to find them. I made a new version of the PHP script, polling every 5 seconds. It also uses two models which try to predict the performance loss. So we can compare the actual performance loss matches with the model and see if it matches. I discovered that recovery is not perfectly linear: recovery per minute is 9/boatspeed%, but boatspeed is increasing as performance is increasing, so its not a constant rate of recovery. In the image, m1_eff is using the assumption that the boatspeed used to calculate the rate of recovery is the actual boatspeed, m2_eff is using the assumption that the rate of recovery is based on the boatspeed when perf would be 100%. I included the raw data, but it's cut off because I has some internet problems. So I'll try again, with some more different manoeuvres, another time. The raw data is a csv file, which can be opened with excel, or another spreadsheet program, but also with a text editor. I included the angles in degrees in the file. I'm looking forward to the day captains order their helmsman to steer ".2 tau" (.2 being in base 12, and tau being 2pi), and pilotes are "requesting takeoff from runway .9 tau" EDIT: the graph (image) has a wrong scale on the x-axis (horizontal), the numbers are seconds, not minutes. --- Last Edited by kroppyer at 2013-10-14 23:39:32 --- Attachments* perf-model-3.png (33.5 KB) perflog_1.csv (9.2 KB)
Posted by javakeda Posts 71 Registered at 2013-03-01 03:18:22	2013-10-15 00:52:38 - L
	Kroppyer: I remain in awe, and am watching and learning. I think the last time I programmed in PHP the version level was 3-something. I'd love to update my version -- just to understand what you are doing. Again, the last time I programmed and ran PHP I used a standalone UNIX server. The application was paired with a MySQL database. I was helping a non-profit group manage their event attendance. This was volunteer work. Not sure what I would need to run your script. I would hope that I could run it in a standalone environment [UNIX, or whatever]. Your knowledge here is leagues ahead of mine. Would love to discuss this off-line. Not sure how that happens. netwrangler@yahoo.com is an email path that might work. Regards, Javakeda Bob
Posted by A2 Posts 347 Registered at 2009-10-26 00:39:42	2013-10-15 11:12:00 - Re: Performance loss
	Posted by javakeda 2013-10-14 17:17:35 Sail Management "Sailing my SOLTP52 in ~17kts TWS earlier today, I reached off from a TWA of -36 to a TWA of -153. I was greeted with an initial PERF hit down to 91.83%" =========================== 117 degrees equals 2.042035 radians divided by 25 equals 0.0816814 Efficiency of 91.832% --- Last Edited by A2R at 2013-10-15 11:52:55 --- None so blind
Posted by A2 Posts 347 Registered at 2009-10-26 00:39:42	2013-10-15 12:17:15 - Rod's Konstant
	For every Degree of course change, that does not induce a Tack or Gybe, there is a .06981317% Performance Loss. None so blind
Posted by sol Posts 165 Registered at 2008-08-14 11:08:14	2013-10-15 12:46:11 - Re: Performance loss
	oops, wrong account - kroppyer --- Last Edited by sol at 2013-10-15 12:48:19 ---

Page: First Previous 2 3 4 5 6 7 8 Next Last

Please login to post a reply.

Races

Next Race: 00d 00h 00m

Current Races:

Garagiste's Test Lab

The first TIMED race of the northern hemisphere Summer takes us to the Gulf of Finland on a 165nm zig-zagging course beginning and ending in the seacoast village of Haukilahti and circling the island of Suursaari. It’s only appropriate that the boat for this race is Finland’s own Club Swan 50. This is a TIMEDrace so you may RE-REGISTER HEREto try again after finishing a run. You will have 13 days and 11 hours to test your skill and decision making after the race opens.
Race #1928
INFOby brainaid.de
PARTICULARS
WX Updates:
0430 / 1030 / 1630 / 2230
Ranking:
TRQ2 - TRCH - SUPSOL - SYC
RACE CLOSE: Saturday,
26 July at 23:00 UTC
Race starts: Jul 13th 12:00 Registration will open soon

Facebook

Login

Support Sailonline

Races

SYC Ranking

Series

Mobile Client