-
Welcome to Overclockers Forums! Join us to reply in threads, receive reduced ads, and to customize your site experience!
Two new articles by BillA on frontpage!
- Thread starter Tecumseh
- Start date
The previous "Modified Swiftech "article* is the one which gives me playable :
The nozzle IDs in Bills graph are 7.061mm (0.278") and 8.738mm (0.344").
Even here I can only represent** my pseudo-science predictions for a 60x60 8mm based(is 462-B 60x60x8mm ?) Swiftech462 style wb.:
Any chance of some results in Excel so can include in plots without resorting to visual inspection of GIFs?
Producton of "T/W" v PD prediction is even more psuedo-scientific.However plotting "T/W" v "Predicted PD from Nozzles in a 12.5mm tube" may possible suggests any trend. Using http://www.pressure-drop.com/index.html for PD in nozzles ( 2 nozzles in series for 3mm case) get
and
* http://www.overclockers.com/articles688/
** Shown,amongst others, here http://forum.oc-forums.com/vb/showthread.php?s=&threadid=80618&perpage=30&pagenumber=5
The nozzle IDs in Bills graph are 7.061mm (0.278") and 8.738mm (0.344").
Even here I can only represent** my pseudo-science predictions for a 60x60 8mm based(is 462-B 60x60x8mm ?) Swiftech462 style wb.:
Any chance of some results in Excel so can include in plots without resorting to visual inspection of GIFs?
Producton of "T/W" v PD prediction is even more psuedo-scientific.However plotting "T/W" v "Predicted PD from Nozzles in a 12.5mm tube" may possible suggests any trend. Using http://www.pressure-drop.com/index.html for PD in nozzles ( 2 nozzles in series for 3mm case) get
and
* http://www.overclockers.com/articles688/
** Shown,amongst others, here http://forum.oc-forums.com/vb/showthread.php?s=&threadid=80618&perpage=30&pagenumber=5
Last edited:
- Thread Starter
- #3
Les, it's not pseudo-science to use established flow tables
and calculations to make projections. That's engineering, right?
The German pressure-drop.com site seems more like somebody's
homework assignment ran amok. Without documentation of
exactly what the code does, how can we be sure of results?
In short, where are the code's validation studies?
The plots seem in the right order for nozzle size, but
clearly there is a off-set issue. Is it because of the code
or the "model?" That's a rhetorical question.
and calculations to make projections. That's engineering, right?
The German pressure-drop.com site seems more like somebody's
homework assignment ran amok. Without documentation of
exactly what the code does, how can we be sure of results?
In short, where are the code's validation studies?
The plots seem in the right order for nozzle size, but
clearly there is a off-set issue. Is it because of the code
or the "model?" That's a rhetorical question.
Last edited:
Thanks Bill .Received 4.34pm.
Makes life easier.
Tecumseh.
Dunno about validation studies, but think will be better than using PD proportional to flow-rate to power 1.75 relations. Which have been using previously. Difficulty to find specific nozzle data or relationships on the web which will apply to the diameters and flow rates encountered. Even with this calculator had to make dubious extrapulations.
Yes, the cooling v Pd predictions I do take with a multiple pinch of salt.
Makes life easier.
Tecumseh.
Dunno about validation studies, but think will be better than using PD proportional to flow-rate to power 1.75 relations. Which have been using previously. Difficulty to find specific nozzle data or relationships on the web which will apply to the diameters and flow rates encountered. Even with this calculator had to make dubious extrapulations.
Yes, the cooling v Pd predictions I do take with a multiple pinch of salt.
Have put experimental data and predictions on single plots.
Have converted my original "T/C" to "C/W" by allowing 0.1c/w for TIM*
Possibly there is some correspondence in the influence of "Nozzle ID" on properties.
Maybe worth and possible (if Predictions 3mm PD v Flow have any credence) loking at the 3mm ID case.
* C/W(TIM) ~0.1 - from Bill Adams work(e.g http://www.overclockers.com/articles654/index02.asp ) ,making an allowance of 0.05c/w for die thermocouple offset.
Have converted my original "T/C" to "C/W" by allowing 0.1c/w for TIM*
Possibly there is some correspondence in the influence of "Nozzle ID" on properties.
Maybe worth and possible (if Predictions 3mm PD v Flow have any credence) loking at the 3mm ID case.
* C/W(TIM) ~0.1 - from Bill Adams work(e.g http://www.overclockers.com/articles654/index02.asp ) ,making an allowance of 0.05c/w for die thermocouple offset.
- Joined
- Dec 1, 2002
Hey Les56. Help me out here for a sec. I'm confused again (happens to me all the time). Higher the CW the better or lower the better?
Higher the CW the better or lower the better?
- Joined
- Dec 1, 2002
Les56 said:
Thanks. Just found out myself too. So it looks the nozzle ID has some benefits for the better. Looks like when you get under lets say 1 LPM that no longer is true and the bigger nozzle works better. Does the normal pump we use do 2lpm with a 3mm nozzle?
- Joined
- Jun 8, 2002
- Location
- Melbourne, Australia
BillA said:
I know that you can buy vane-based fuel pumps relatively cheaply, but they are pretty low flow. Two of those in parallel should get you up around 15lpm with some pretty high pressure, all running off 12v. Apparantly a bit noisy, and I don't know what the lifespan effect would be with running water as opposed to petroleum.
Do you know of any good cheap(ish) gear pumps Bill?
Last edited:
- Joined
- Dec 1, 2002
BillA said:
I like to see the results of that Bill. With my current hydor I don't see much of a change with a smaller nozzle and this is pretty much with a flat base although tapered a bit from the center. If I run lets say a pump with 50 PSI of pressure and the same flow as the hydor, the temp changes are unbelievable. Didn't effect idle temps much but load temps where just about the same as idle but maybe 1C higher.
- Joined
- Jun 20, 2002
- Location
- Winnipeg
The "pressure drop vs flow" graph shows the 6mm inlet (red) flowing better than the 7.04 inlet (yellow). Shouldn't it be the other way around?
What happens when two blocks in the same system have the same c/w but have different flow rates? Won't the radiator water temp (relative to room temp) be affected by the different flow rates?
Even though a high flowing waterblock may have a worse c/w at the same pressure drop, it should have a better equilibrium temp due to the improved rad performance. In some cases this may offset the c/w performance enough to beat a block with a better c/w. Since the "c/w vs pressure drop" graph can't show this, you can't conclude which block is better in any system.
What happens when two blocks in the same system have the same c/w but have different flow rates? Won't the radiator water temp (relative to room temp) be affected by the different flow rates?
Even though a high flowing waterblock may have a worse c/w at the same pressure drop, it should have a better equilibrium temp due to the improved rad performance. In some cases this may offset the c/w performance enough to beat a block with a better c/w. Since the "c/w vs pressure drop" graph can't show this, you can't conclude which block is better in any system.
except that the effect of the flow rate upon the rad is MUCH less than for the wb,
the wb effect will dominate the system performance
you should focus on the pump
be cool
Edit: forgot Cathar's question;
Iwaki gear pumps come up on eBay every once in a while
WAY too expensive new, $650 by specail order from Japan
and the US made ones list at over a grand !
Procon vane pumps (4gpm @ 250psi) take a 1/3 hp motor, ok for testing but not practical for actual use
-> better to optimize the design for a pump, than to seek a pump to push whatever
the wb effect will dominate the system performance
you should focus on the pump
be cool
Edit: forgot Cathar's question;
Iwaki gear pumps come up on eBay every once in a while
WAY too expensive new, $650 by specail order from Japan
and the US made ones list at over a grand !
Procon vane pumps (4gpm @ 250psi) take a 1/3 hp motor, ok for testing but not practical for actual use
-> better to optimize the design for a pump, than to seek a pump to push whatever
Last edited:
- Joined
- Oct 12, 2001
- Location
- Los Angeles
BillA (or anyone who knows), refresh me please; What is the temp difference in degrees for every drop in C/W?
take that "C/W" value and multiply it by the CPU generated heat,
this will yield the temp rise
note that there are 2 rather 'hairy' assumptions (which is why, in part, "C/W"s are not comparable):
- how accuratly was the 'heat' characterized, and
- how controlled/consistant is the TIM joint temp offset
saying nothing at all about the temp measurement accuracy, eh
be cool
this will yield the temp rise
note that there are 2 rather 'hairy' assumptions (which is why, in part, "C/W"s are not comparable):
- how accuratly was the 'heat' characterized, and
- how controlled/consistant is the TIM joint temp offset
saying nothing at all about the temp measurement accuracy, eh
be cool
Freeloader said:
You're comparing measured test results to a simulator's results. Yes the 7.04 inlet should permit a higher flowrate for the same pressure drop. The most probable cause for this 'reversal' is that the simulation is lacking in some respect.
Les,
Does the simulation take into account all the internal dimensions of the waterblock, or just deal with impingement on an infinite flat plate?
Freeloader said:
I understand your question as:
'If I have a system, (block X, pump, rad, tubing, etc.) and I swap the block for new block Y without changing anything else, and I find that block Y has the same C/W as block X did when run in the system, but the flowrate is higher in the system with block Y, won't the performance be better because of increased radiator performance?'
I'd answer that question with 'probably'. Some of the bizarre curves Bill has measured for radiators show that some radiators can actually decrease performance with increasing flowrate over a certain range of flowrates. This is not the typical case though.
Another issue is pump performance. Centrifugal pumps draw more power when they are moving more water. With a submerged pump particularly, the increased power consumption of the pump might offset a benefit in the rad due to flowrate.
Freeloader said:
You may not be able to conclude which block is best for ALL systems. You can conclude which block is best for use with ANY SPECIFIC SYSTEM. (Provided you know the relevant characteristics for all the non-block system components.)
I expect the White Water data to show that (at least for the time being) there is one best commercial block for all systems. With the data we have to date though, designing an optimal system is a very complicated process.
- Joined
- Oct 12, 2001
- Location
- Los Angeles
OK well I mean in your tests, what are the 'estimated' temp differences per C/W? If block 'A' has a C/W of 20 at 8lpm and block 'B' has a C/W of 19 at 8lpm, what difference in degrees would that equate to?BillA said:
yes, quite soSince87 said:
but not really that difficult if one starts with a specific pump's selection, then choose the other components accordingly
- given the somewhat limited pump choices (at a 'reasonable' cost), the process simplifies itself rather quickly
WW up tomorrow (so I'm told)
be cool
EDIT: jeez nikhsub 1, did you not read it ?
70W actual, ~100W per Radiate
so multiplying by 100 = ?
2ed EDIT: ok, I see what's happening
JoeC split the summary off of the Multi-Hole writeup,
so the applied power description is now not included in the summary
- hop back and read the other one
Last edited:
BillA said:
Yes, picking some piece of hardare as a starting point makes the rest vastly easier. I'd have guessed that the block should be chosen as the starting point. After all, how pretty the block looks is the most important issue, right?
