Water pump impeller design & pumping efficiency

[JJDraper]

I have just noticed a major difference between my old water pump and the new one. The impeller design is very different - see the pictures. The new one is very crude in terms of the impeller blades compared to the old one, although closer to the original (looking at the workshop manual). The reason for changing the pump was a failure of the long sleeve bearing after many miles, but the engine has never suffered from overheating.

The difference between the designs is striking, with curved blades on the old impeller, compared to flat/straight blades in the new one. This is important because curved blades improve the efficiency of pumping (in terms of volume per revolution) significantly. Technically, flat bladed impellers are more suited to mixing applications, with curved impellers suitable for pumping applications.

Does anyone have any ideas why the efficiency of TC water pumps is hobbled in this way? I for one would prefer more circulation of coolant!

Jeremy

[elansprint]

Jeremy who supplied the impellor ? I seem to remember when i rebuilt my engine a long while ago Mick Miller saying that some suppliers had impellors with less blades on that were less efficient
Ian

[Donels]

I don?t think your point about curved blades being more efficient is correct. If you look at your original impeller the water flows from the centre outwards, so the flow is governed by the throat area between the vanes near the centre. The flow cannot increase as the vane area increases as water is incompressible, so cavitation occurs downstream of the throat. Hence this design is fundamentally inefficient.

That later design by comparison is much more open and allows easier circulation, so is probably much more efficient, ie more flow per revolution.

[promotor]

Although the comment I'm about to make doesn't relate to efficiency between the two designs of impellor in your post, I would recommend that you double check how much clearance each vane gives between it and the sealing ring / o-ring carrier.
The one I bought recently (similar looking to your new one in the picture) was part of a rebuild kit and wasn't machined accurately at all - it gave a variation in clearance measured at the vane of over 0.030" from tightest to largest once installed in the o-ring carrier.
I thought that was pretty poor so I modified it until all of the clearances were similar with the tightest ones being to the clearance spec given in the manual.

Hopefully mine was the exception and not the rule, but worth checking all the same.

[USA64]

It might be the speed it was designed to turn, racing vs street? I know boat propellers gets very complex, pitch, diameter, speed etc.

[el-saturn]

i also dont like the simplified "wings".... NOR do i like the rusty one. my understanding of hydrodynamics is limited BUT i also remember studying a a very complex "ROLLA offshore boat propeller" - evidently he's the world's best ? which we intended to make in cfrp ------------- i then even talked to stefano casiraghi at a race in virareggio to see if gancia; his winery promoter would pay for the development of a carbon propeller --- they didn't! you should here the sound of a 20 boats (@ 4500hp per unit = 90'000hp!) START: most of em were 8L V12 Lamborghinis (some seatec diesels 3x1500hp) ------ for greta thunberg (GT !! our new anti-green LETTERS!!!) pure hell - for us the contrary! sandy ...................... and no, THEY didnt give me a ride, HOWEVER the next day i DID get a ride in a SEATEC offshore boat (4500hp) on the Lake of Como: he only went to 80 knots AND the day after i visited Lamborghini where they had one of those 8L V12s on the dyno - an engineer asked me if i'd like to hear it: i did ------- makes a street lambo sound vegan!! GT

[RichardHawkins]

Jeremy,

I cannot be certain as to which of the impellers is original, only that my 1968 Elan had the straight blade type when I bought it in 1979.

As a general statement, curved vane impellers are more efficient than straight. However, a centrifugal pump usually has the impeller situated in a volute casing, and the smaller the clearance between the casing and the impeller the more efficient the pump. Our impellers are just in the block, so I would expect little noticeable difference in efficiency.

Hope this helps,

Richard Hawkins

[types26/36]

That?s Interesting, I?ve never seen one of those impellers with 8 curved blades , was that fitted on early cars or was it a special application for racing as mentioned? the hole for the shaft also looks to be a large diameter or it just how the camera is portraying it?
Here are a few I?ve changed over the years, the top left is a hard black plastic impeller and the one next to it is alloy.
The rest are all cast iron although I have seen one that had a thin steel impeller but all had 6 blades and not 8.
???and yes! I know I should have thrown them all away

[rgh0]

Jeremy,

I cannot be certain as to which of the impellers is original, only that my 1968 Elan had the straight blade type when I bought it in 1979.

As a general statement, curved vane impellers are more efficient than straight. However, a centrifugal pump usually has the impeller situated in a volute casing, and the smaller the clearance between the casing and the impeller the more efficient the pump. Our impellers are just in the block, so I would expect little noticeable difference in efficiency.

Hope this helps,

Richard Hawkins

+1 to comments above. Generally curved bladed pumps are more efficient but many other factors are important. Given the crude design of engine water pumps it probably makes little practical difference. I have never seen a curve bladed pump impellor in a twin cam and given the impellor shape not sure how well it would have fitted in the casing.

cheers
Rohan

[StressCraxx]

Pumps and seals are my day job. This radial, semi-open impeller is perfectly adequate, especially considering there is no discharge volute in the engine block.

[Quart Meg Miles]

In fifty years I've only ever seen the simple 6 blade metal type but I've never thought to measure the blade height.

[JJDraper]

[quote???and yes! I know I should have thrown them all away ][/quote]

I am the same - keep all my old head gaskets and other old parts etc as they each tell a story of why they failed or wore out.

Getting back to impellers - I should mention that this pump had been in service for well over 100,000 UK miles without incident, indeed the engine has been very well cooled, hence my remark that the old impeller must have been efficient. The shaft is identical to normal (the photo makes it look bigger), but the narrower boss diameter gives a larger inlet throat.

The clearance to the housing was close, evidenced by the clean vane edges, but I have no measurements and the old housing is now built up with a new QED pump kit. The design and manufacture of a part such as a cast iron impeller is not trivial, so we come back to why my old unit appears to be more 'engineered' than the new ones? The blades are gently curved and streamlined in a way that suggests either serious design or someone with an artistic bent!

Looking at the aftermarket for auto water pumps in the US of A, there is a lot of talk about 'more efficient designs compared to mass market fitments'. Is this just hooey to sell stuff? Perhaps the comments in this thread about different designs for different rev ranges is important. Cooling is an Achilles heel of the TC engine, so I am surprised that there isn't more effort to improve the coolant flow around the engine. Most effort seems to go into making changing the pump easier, than making it more efficient.

Given that this item was clearly well made and must have been produced in numbers, I am surprised that no-one seems to have ever seen one like it before. If I start getting cooling problems, I may get the old impeller cleaned and refit it!

Jeremy

[gus]

The 'new' one is a normal Lotus/Ford impeller

The 'old' one is something either someone made or crossed from another vehicle.

Would be interesting to do some testing.

I would not worry.

A relatively stock Twink with a stock radiator, good thermostat and an electric fan will not overheat,so I don't think the pump is a problem.

[JJDraper]

Browsing the archives (having the in-laws round for Christmas motivates me here) I realised I missed this thread a few years ago:

viewtopic.php?f=39&t=22662

The two threads make some interesting reading.. Still no conclusions about the impeller design, except that it looks like Burtons tried a fancy one a few years ago. My continuing research on pumps and impeller design suggest that that there is room for improvement, even to the extent that the block cavity does work as a crude volute for the pump outlet. I also found this cross section view of the water pump assembly - if any one has a better one, please post it as it greatly helps understand what is going on in the pump.

Jeremy

[el-saturn]

jeremy - at one time in my life i'd be drawing engine components: from diesel engines all the way to bits for the megatron 1500cc turbo engine (= BMW F1 engine .... piquet) ---------------- THE MOST IMPORTANT bit in pic. 4 is missing: the f... ceramic "washer" !*"+!?`!!????$$*+? WHY? or is it my eyes?