MG-Cars.info

Welcome to our Site for MG, Triumph and Austin-Healey Car Information.

Parts

MG parts spares and accessories are available for MG T Series (TA, MG TB, MG TC, MG TD, MG TF), Magnette, MGA, Twin cam, MGB, MGBGT, MGC, MGC GT, MG Midget, Sprite and other MG models from British car spares company LBCarCo.

MG TD TF 1500 - Results of flow tests on 3 impellers

Here finally are the results of the flow tests I did on the 3 impellers, the standard 4 vane impeller that came with all TDs and TFs, the 6 vane impeller currently being used by most rebuild services when they offer an upgrade, and the newly designed 6 vane impeller that I had 3D printed on shapeways.com, using a design with a narrower rake and a curved blade.

I built a testbed on an outside table, using my drill press motor and a jig which held the water pump. I mounted a 100 gallon stock tank on the table and filled it. The reason for this was to avoid any vagaries of water pressure that might occur if I simply connected a water hose to the pump; Tucson water pressure varies from about 30psi to 80 psi. With 100 gallons in the stock tank, I had a reasonable "head" of pressure that would not vary significantly over the course of the test (as we will only be removing 7 gallons from the tank each time). Before each test I added water to bring it back up to the same level in the big tank.
I then took a small galvanized tub and using a gallon bottle, measured how many gallons it took to fill to a known point on the tub. It took exactly seven gallons to bring the water up to the top of the first "lip" of the tub. Hence this is now a 7 gallon test...


I connected up the pump and the tank, opened the tank, and held the output hose from the water pump out and down, until water began flowing without the pump running (water can move easily around the impeller and out). I then raised the output hose until the water stopped flowing by gravity, then held it at that point for the rest of the test.
Then I started the pump and the timer simultaneously, and waited until the pump had filled the lower galvanized tub to the 7 gallon level. I stopped the timer and turned off the pump.
Here are the times it took for each impeller to transfer 7 gallons of water:
Original 4 vane water pump: 8 minutes, 16.21 seconds
Standard 6 vane water pump: 5 minutes, 48.81 seconds (nearly a 1/3 improvement, not surprising with 6 versus 4 vanes this seems pretty appropriate!)
Modern 6 vane impeller pump: 5 minutes 18.7 seconds.

The curved modern impeller with a narrower section definitely moves more water, approximately ten percent more.

Beyond this, I make no claims for the impeller, and clearly a standard six vane, if you can buy it, is a good bang for the buck. But the modern impeller offers a 40 percent improvement over the traditional 4 vane, vs about 30% for the standard 6 vane currently on the market.

For pictures and links, look for the original threads on the improved 6 vane impeller.
Here is the shapeways link:
https://www.shapeways.com/product/5Z7278JJ2/mg-td-tf-improved-water-pump-impeller?li=shop-inventory&optionId=56452364


I hope this data proves useful to people when they are looking at working on their cooling systems.
Geoffrey M Baker

Nice job, Geoffrey! That's interesting and valuable data.
Steve Simmons

Geoffrey,
Could you explain in non-engineers terms why I would want to buy one of these for $110? My TD runs at normal temperatures and even a tad cooler with the MGB 7 bladed fan which I installed after my old blade snapped of. Curious.
Thanks,
Ed
efh Haskell

Geoff,

Well done, that's what I call going beyond the call of duty. I think John James should put this in the next issueof TTT2 and I have sent it to him. Maybe Bud can add it to T-Talk.

Jim
James Neel

Ed, I do NOT recommend any particular solution. I was interested in a theoretical idea (there was a design for a modern 6 vane impeller with a better blade design available on the web, and we took that and improved it), and to build a real model using 3D printing and see what improvement it offered. That was my only goal.
We have now determined that it offers a small but significant improvement over the existing 6 vane impeller on the market - but at considerably more cost.
So I don't recommend it over any other solution, but there is at least one scenario where it might be useful. To anyone wanting to upgrade their own pump, they could install this improved 6 vane impeller for an approx. 40% improvement. As the standard 6 vane impeller currently on the market costs $50 and requires complete machining (it's merely a rough cast), the finished 6 vane version on shapeways at $110 seems relatively competitive to me (assuming hourly rates at a machine shop to machine the standard 6 vane impeller from it's rough original cast up to working condition.)
Now, you bring up other cooling alternatives. I'm not knocking them or trying to compete with them, but here's my 2 cents...
For anyone not wishing to change the look of their engine by switching to a non-original fan, the 6 vane impeller seems to offer a similar cooling improvement for a reasonable cost.
For anyone with heating issues, this might be another tool in the toolbox for keeping their engine temperatures down, in association with other improvements such as a MGB fan, or an electric fan, or a recored or rodded radiator.

Lastly, if what you are arguing for is the best possible improvement for the best possible cost, I suggest you read my "Rodding a 51 MG radiator" thread for a technique to completely clean out your radiator and bring it back to original factory specs, at zero cost (well, maybe $5 worth of solder?). This should improve the cooling in any system, well beyond efficiency offered by either an improved impeller OR an improved engine fan...
Geoffrey M Baker

Geoff, the results are indeed very interesting. You might want to re-investigate your option prices. The last I knew, Butch Taras only charges $35 for a bronze, 6-blade impeller when rebuilding a water pump. Bud
Bud Krueger

Bud, the reason is that the bronze impellers are made cast in batches, probably by the thousand - hence much lower prices.
I have no interest in trying to make 1000 of these improved versions; so the cost will not be competitive with the regular ones.
However, you cannot buy the standard six vane impeller (I know of no-one who sells them individually) except as un-machined pieces.
Someone rebuilding their own pump can install the shapeways impeller for less than what it will cost to buy and machine their own impeller off ebay, if they do not need a full pump rebuild.
That's really the only "market" I see for this impeller, until someone wants to invest in having a batch made by casting. I don't have those deep pockets!
But the important thing is the data is out there for everyone to read, and hopefully help in making their own decisions when redoing their cooling systems...

The problem with doing a pump rebuild, of course, is the issue of modern seals. There is currently a new seal that fits, sold by Mr Schapel in Australia, and I saw recently that someone in Germany had acquired some TF type seals and had them for sale. So doing a rebuild without having to rebore or machine the water pump is still a possibility.
Geoffrey M Baker

Geoffrey,

The flow of the radiator has to be matched by the pump and visa versa.

You risk pumping the water that won't get crammed through the radiator out the overflow pipe. Open systems will do that.

....
MAndrus


Geoffrey,

You have done some very good work here.

Can you tell me the ID of the output hose that you used? The one that goes into the water collection tank.

If you are up to it, I would like to suggest an additional Investigation.

First the distance from the center of the TD pump to the bottom of the radiator tank is about a foot.
Your test was done at Zero head. In fact there is a 1 foot head.
I dont expect this will make a difference since one foot is only about 0.6 psi, but it would be interesting.

Second you said you connected the drill press (DP) motor to the pump. I checked both my DP motors and they are 1750 rpm.

I think that a concern would be cooling in slow moving traffic.
I note that the pump pulley seems to be equal to or just a tad smaller than the crankshaft pulley.
I wonder if you could check the flow, with a bit of a head at around 1000 rpm.

Again very nice work.

Jim B.
JA Benjamin

Jim, I've already torn down the test jig (I needed to reassemble my drill press :) )...
The purpose of the test was NOT to give a benchmark for the impellers and their output in a way which in any way matched the cooling system setup in the car, but simply to give an accurate comparison between the three impellers. There are a billion factors which will vary the output results (hose diameter, jig setup, pulley size, belt size, belt age, belt torsion etc etc etc). All I wanted to do was get a fair comparison test between the impellers, which I succeeded in doing. For future reference, anyone can now assume that the standard six vane is approx. 30% more efficient than the 4 vane, and the modernized six vane is approximately 40% more efficient.
My output hose was a standard garden hose mounted on a 5/8 OD tubing brazed to the mounting jig.
My mounting jig was just a piece of large OD tube (2 1/4"?) which fit over the water pump mount, which I welded shut on the other end with a flat plate. Then I drilled a 5/8 hole for the output tube, and brazed that on. It does not resemble, or match in any way, the engine chamber where the water pump mounts, so the actual flow rates will be entirely different when mounted on your engine. However, the percentage difference of flow between the impellers should be identical on any engine with any setup.
I believe my speed was 1300 rpm, using the middle pulley on my drill press motor.
As far as cooling in slow traffic, well that's why I'm doing all this :) ... my engine overheated in a parade. While I await it's triumphal return from the machine shop, I'm rebuilding the water pump, radiator, etc etc!
What I have said (in previous threads) is that if - as people generally report - the standard 6 vane impeller helps reducing engine overheating, this new impeller will improve on that a bit more. That's all.

A lot of people seem to think that either their approach is better (different fans, recored radiators, electric fans, waterless systems, etc etc) or that any additional improvements are simply unnecessary ("my car runs fine ever since I added xxx").
My point is simple ... here's another tool for anyone looking to improve their cooling system. That's all it is. If you want to go another route, good on yer, mate!

Mandrus, in regard to your point; I believe we are a long way from moving too MUCH water through the system. As everyone has reported positively on the existing standard 6 vane impeller, my new impeller which is 10% more efficient should be even better, in theory. That's the only claim I am prepared to make. I think you'd need a much higher volume pump before you reached a point where efficiency ceases to increase and actually drops. Yes, in racing engines, they do have this problem, because they are eternally "pushing the envelope". In the MG world, our envelope is pretty small, our pump is tiny, our water pressure is pretty low, and we have plenty of room to push.
Geoffrey M Baker

Geoffrey,

To REALLY get serious, we'd need to install each variation of pump in a car fitted with an accurate electronic sensor and test for an actual change in temperature. It could easily be done - same driver/day/route/coolant mix, etc. Also, a static bench test with 95*C water to calculate the added load vs the factory pump will tell us what the the extra work will cost HP-wise at what rpm range.

I'd give it a go. Really. Send one over.

40 years ago a "better" water pump was offered for Model A Fords. The consensus was that pumping the extra water would cause more heat to be generated and rob us of some of our precious 40 HP.
MAndrus

Well now you've come up with a whole new argument for improved impellers... Use a bigger pulley, fewer rpm, less energy needed, 40% fewer hp robbed from the engine, while moving the same amount of water as the original... 😃
Geoffrey M Baker

GMB, I don't think there's a free ride to be found here. Moving a given amount of water takes the same amount of power no matter how fast or slow the rpm's. Lifting a gallon of water 1 foot takes the same amount of energy no matter the size of the container be it a teaspoon or gallon jug. Both require the same ooomph. It's a law.

Power lost to a design that just flails the water around the chamber instead of lifting it is a totally different story, so a redesign of the blades and pump chamber might be required, but that isn't going to happen without some serious computer work. This is why you need to do a test that measures the energy required to move the same amount of water with each of the pump designs.

Each blade generates friction and needs it's own amount of power. 6 blades will require 2x the power of 3 blades but will not move 2x the water because the extra blades take up a measurable amount of volume and reduce the amount of water that the propeller would otherwise have had the opportunity to move causing the need to turn further, blade efficiency and some other physics stuff not withstanding.

Putting an ammeter on the cord would have shown the change in power input.

I'd still like to test a few if you get that far. Our cars were designed with 1940's technology... and slide rules. Frankly, a redesigned more efficient blade design rather than more of them will probably be the better mouse trap you're looking for.

.....
MAndrus

Geoffrey, I hope you don't feel that I'm jabbing sticks in your cage. I'm just joining in on the thought process and sharing what I've been taught and experienced.
MAndrus

My TF had a brand new radiator fitted in 1968. I have a receipt showing it cost £15. However completion of the rebuild way delayed for a short time and it was not filled with water until last year! Just to be sure I took the radiator, heater and petrol tank to a firm in Glasgow for testing. They found the radiator, even though unused, to be unserviceable. It was heavily furred inside. Apparently if it had contained water and antifreeze it would have been OK. Dampness had ruined it. The purpose of this response is to explain I had to have a new core fitted at a cost of £120. The repair company explained the new core is much more efficient than an unused old one. Isn't it better to upgrade the core rather than modify the pump?


Jan T
J Targosz

Jan, people seem to get caught up in an "either/or" approach when talking about these sorts of issues. Is it better to upgrade the core than modify the pump? I would think that depends. When talking about recoring radiators, I've heard people talk in the $300-$400 range. So if you have the money, that's a good option. For someone more budget conscious, a 6 vane impeller instead of a 4 vane is a better deal that should result in cooler running temperatures (as reported by the vast majority of people on this forum who have installed rebuilt pumps with 6 vane impellers). For those not worried about 'concours' looks, an MGB fan is even cheaper. Cheapest yet - what I did with my radiator, which was to "rod" it myself, desoldering it and then cleaning it from the inside; followed by reassembly and carefully bending the fins back wherever I could - all at the cost of a few bucks of solder. I would guess that the outer rows of the tubes were severely occluded, and even the cleaner ones still produced plenty of dust and particulates when "rodded". How much more efficient is my system now? Between rebuilding the engine and cleaning out the water passages (which were totally occluded, according to the machine shop) and rodding the radiator, I've probably brought the system back very close to original specs, and expect to find myself with a much cooler system overall when everything is put back together.
But should one recore the radiator or replace the pump or pull the engine to clean the passages... or all of the above?
There is no single answer. There's just a range of options, which people will use depending on circumstances and budget. (maybe they just don't want to pull their engine quite yet to clean the passages, and can get away for a few more years with an upgraded pump or radiator...)
In my case, I'm glad I had to pull the engine, now I KNOW what issues I was facing. Until you've thoroughly examined EVERY part of your cooling system, you really don't know where the problems lie. A new radiator may just hide the problem for a few years, while a blocked engine block continues to obstruct water flow, keep engine temps high, and increase the wear and damage being done...
So my advice is check EVERY part!

MAndrus, I don't mind at all. Your statement that "Moving a given amount of water takes the same amount of power no matter how fast or slow the rpm's" is of course true, in an ideal world. But every part of a system is always less than 100 percent efficient, so when you improve one part's efficiency, you improve the system. My 6 blade impeller turns at the same rate as the 4 blade, but it moves more water, thus improving cooling. Does the increased amount of water moved rob the engine of hp? Of course it does but is it measurable compared to the amount already being stolen by a primitive 4 vane impeller, whose thick blades, rough cast, and poor design already impede its own efficiency? I would doubt it. The pulley turns at the same speed, but more water is moved. I would be willing to bet that the amount of engine hp lost is simply insignificant, between the two impellers. Or I would imagine we would have heard about it, from the hundreds of users who are already using 6 vane impellers. "Yes, my engine runs cooler, but it doesn't have the same pep...?" Haven't heard that, at all - and people have been using improved impellers for many years, now.

As far as the difference between my 6 vane and the existing 6 vane; energy wise, while 10 percent more water is being moved, I very much doubt any measurable loss of hp can be detected - whatsoever - because most of the efficiency comes from using thinner blades, and curved to boot. The result is the impeller spends less time fighting itself, and the thinner blades allow more water flow than the existing 6 vane impeller, without in any way changing any other variables in the equation. Spin rate is the same (and the impeller weighs less, by the way) so I think while you may argue there is an hp loss from a 4 to 6 vane impeller, I doubt there is any between the two 6 vane designs; mine is just more efficient.
Geoffrey M Baker

Slowing the water pump pulley will slow the cooling fan. Not the best idea. The engineers who designed these cars were very bright guys. They could have used a larger pump with a more powerful impeller or even a different fan. What they used was adequate for most conditions the car might encounter. In very hot climates a boost in cooling can help of course, but most cooling issues are not due to the basic design. Age, corrosion, state of tune and changes in design are always the cause.
Steve Simmons

I wasn't being serious about changing the pulley size, Steve. (Hence the smiley on that post!) I was merely making the point that you could reduce parasitic hp drain on the engine by making a more efficient impeller move slower, thus keeping the same water flow rates but using less engine power to do so. As you point out, doing such a thing would reduce the cooling fan as well.
I suppose (in theory) you could offset this by using a better fan (MGB) or an electric fan.
My MG resides in Tucson AZ. We meet the criteria for "hot climate" here :)
When I have my engine back in, with a nice clean radiator interior (rodded) and a clean set of water passages in the block, and an improved water pump, I expect no cooling problems at all. At least, that's my hope!
Geoffrey M Baker

Sorry, the thread has become so long that I was speed reading through it.

A cool-running TC is not a dream, mine has run cool in every climate, even on long 60 MPH grades in 115-degree Nevada heat. You just have to have everything in good condition and tuned properly. Anything out of spec or tune can cause heat buildup.
Steve Simmons

This thread was discussed between 22/05/2015 and 26/05/2015

MG TD TF 1500 index

This thread is from the archives. Join the live MG TD TF 1500 BBS now