BIG BLOCK alternative Motor

Oh no surprises there - I'm a B/Eng Mech/Elec. But surely there is a proper heat path to those side covers - or is there a big gasket that is thermally isolating them?
You mean if I take a laser thermometer to the side cases they will be twenty or even forty degrees cooler than the radial casing???
 
nope but as soon as you start sinking heat there your will measure your 20K temp difference. heat path is simply too small.

But surely there is a proper heat path to those side covers
Click to expand...
no. why should it?? fins are mount radially.. so? 90% of the sinking surface is radially.. why should the constructor want to design unnecessary heat path there? it is the furthest available surface from the coils available so the least desirable?
 
Because it attaches to the side sheets of my battery case/frame. The radial surfaces are just a conduit.
 
Samd said:
The radial surfaces are just a conduit.
Click to expand...

Um, don't understand what you are getting at. The radial surfaces are the heat sink. The side plates will be significantly cooler than the radial surfaces. Adding heat sinking capacity to the side plates will have minimal returns if any.
 
I think I found out why my motor is getting hot. The kelly is peaking at 150+ amps :lol: that is more than 8kw into the motor peak with my worn out lipo (only 17s left alive lol)
With a fresh 18s limited to 100A peak I should be able to keep the motor from dying while I figure out watercooling.

I am thinking about copper tubes that would fit snug between the fins and glued with higly aluminium charged epoxy coumpound. I already have a small 12v pump, and I think a big PC watercooling radiator could work.
 
How about replacing the bolts with hollow ones into the body and pass fluid/oil through them with piping pressed onto the heads? (if the bolt holes go all the way through of course)

http://www.zendistro.com/products03-2021.html
2021-B.jpg
 
One could machine new side plates to captivate a o-ring on a shoulder so a tube could slip over. The side plate would need to be silicone sealed to the extruded aluminum motor housing. I think this would be very effective but a little difficult for some to get done. Only problem is the damn wire lead is notched into the housing. It would need to come out of the end plate and the hole filled and milled. Is there enough room for the wire to bend and come out the side plate?
 

Attachments

  • liquid cool.JPG
    liquid cool.JPG
    81 KB · Views: 2,969
I'm getting at creating a thermal path to get the heat out of the radial case. Because air is a shit thermal conduit.

Don't use the radial surface of the case to radiate heat of sudden bursts. Move a decent amount to the rest of the bike to carry it away as a thermal shockabsorber for sudden bursts. But reluctant to point it out due to your style crossbreak. Aluminium has half the thermal conductivity of copper. And I make aluminium bikes. Or frocking big heatsinks with seats and wheels if that's your bag.
 
Remember the heat is generated in the coils, so the best way is to get it out of there before it burns insulation etc.
Ventilate the motor + a fan assist ?? ..has been successful in various other applications.
Oil bath to move the heat faster to the housing ?

OR.. figure out why so much heat is being generated ??
 
The problem isn't shedding the average amount of heat over a minute or two, or five when cruising.
It's spreading it when you make a 15 or 30 second burst. If you can carry it away to the chassis then a larger surface can store and slowly diffuse it.
Yep, oil baths help get it from the coils to the case. Peanut oil is good, high burn temp, low oxidation.
Hey mom, I'm an e-Wok.

You're basically turning your drive into an electric jug momentarily due to the surges. A cupfull of fluid is going to heat fast. Would need a radiator.

Air and small fans may be undercooking it, if you pardon the pun. If you can make the frame the short term heat sink you aren't adding complexity or weight if the initial design can incorporate it.
 
I see the bigger problem as being able to get the heat away from the coils quickly..before it damages them.
There is a big insulating layer of air between the coils and the case so changes/additions to the case mass is just too late to help the coils. The biggest heat sink in the world is no help if you cannot transfer the heat to it fast enough
A little oil splashing around the coils has proven very effective on hub motors, and a fan assist air flow has also shown itself to be much more effective than anyone believed it could on a Cyclone set up ( searching for the thread link )
Here.. http://endless-sphere.com/forums/viewtopic.php?f=28&t=56666&p=845919&hilit=cyclone+cooling#p845919
 
Samd said:
I'm getting at creating a thermal path to get the heat out of the radial case. Because air is a shit thermal conduit. [...] But reluctant to point it out due to your style crossbreak
Click to expand...

As long as you dont throw you bike into a river, air is the only available heat sink on a bike. The only way too improve cooling is to enlarge the surface and shorten + widen heat paths though.

That is physics and has nothing to do with "my style". "Moving" heat from a 5kg motor to 1kg of frame alloy that is available closely lacks of sense in my eyes.. even if you add another kg of copper for enlarging that heat path. I think the motor itself has more than enough mass to accumulate heat from bursts

Anyway i think that watercooling is something one should start thinking about if we got some dyno figures from the high torque region. before this, we dont know if it is worth it. I guess it is not...at 100A this motor may well get an electric heater. A fan on the back side and a duct may cure heat problems. That is a simple way we can observe at almost any stationary electric motors used in this power region.

Hillhater said:
There is a big insulating layer of air between the coils and the case so changes/additions to the case mass is just too late to help the coils
Click to expand...
As for as I could see, the stator is in direct contact with the radial alloy housing. Where is that air gap?
 
crossbreak said:
Anyway i think that watercooling is something one should start thinking about if we got some dyno figures from the high torque region. before this, we dont know if it is worth it. I guess it is not...at 100A this motor may well get an electric heater. A fan on the back side and a duct may cure heat problems. That is a simple way we can observe at almost any stationary electric motors used in this power region.
Click to expand...

At 100A phase current this "1000w" 5.8kg version of the motor makes 29.8NM of torque and produces about 1720w of heat, real, and measured at the battery -so i spose the kelly etc (running 33khz pwm not 16.6khz) could use around 100w so the motor has a good 1600w of heat to shed.
its still not saturated though, mostly copper loss at 30NM output..

and the halls seem to be very badly placed on these motors, the yellow one in your pic adam, looks like its a fair bit out. :wink:
 
interesting..how can we find out at which current it saturates? We dont know phase resistance since no one took the time to measure it yet... how can we calc copper loss without having this value given?

Edit: we have this value given from the vendor that was posted below the dynograph as "max effiviency" point for 48V. Here phase current should be equal to battery current
Eff. = 87.9% I = 16.17A P = 774.3W

So P_loss = (1 - 0.879) * 774.3 W = 93.7W ... P_loss = I² * R/2 ... so R = 2*P / I² ... R = 2*93.7W / 16.17A² = 0.717 Ohm

This value sounds too high for a motor of this size.. I'm keen to see what we actually measure. However with this value we should get a loss at 100A of P_loss = 100² * 0.717/2 = 3585W
 
From the overall efficiency calc. What are your science qualifications? Any thermodynamics? Elec eng?

Your little fan with air is too small.
Thermal conductivity of air about 0.02 watts/mK
Aluminium is about 200, and copper 400.
My frames have seven kilos about.

Go outside, look at the transformer on the power pole in your street. Its alloy full of oil. Not air because its a bad heat receptor.Nor water - its electrically conductive.

The stator has varnish between wire and alloy case. Doesnt matter if the gap is visible to your eyes or not. Its there.

You said it yourself, enlarge the surface.
 
where do you think the heat goes once it is in the fluid? It gets sunk into the air by a radiator...sure air is a bad heat receptor. But there is lot's of it.

The stator has varnish between wire and alloy case
Click to expand...
varnish is not air. It is a much better heat conductor. also, there can be epoxy between varnish and lamination. Looks like there is none yet.. but we could put it there. However, i have never heard from stock BLDC motors that run their rotor within the fluid.. This really would be something new. Sure the Direct drive people do so with success. At least worth a try.
 
Yeah what about drilling two holes on the rear cover for in/out flow and have an external radiator and pump setup... with some thinking you could even use the motor as an actual pump. The main issue I see is that this method is creating a direct heat path from the windings to the magnets :|
Sealing the motor should not be too hard to do.
Then , what kind of fluid should be used in this application? Water/glycol? antifreeze? atf oil? (I hate oil)
 
transformer fluid sound like worth a try.. we need something non-hazardous and very thin...so it does not produce too much drag.
The gap is l = ~0.5mm. The circumference of the rotor is ~0.25m, the width is 0.05m so the moving surface area A = ~0.0125m²....At 2000rpm the radial rotor surface moves at v = 8.3 m/s or ~30kph past the stator teeth. Sadly i dont have the formula at hand to calc the fluid friction here. Edit: here it is http://wps.aw.com/wps/media/objects/877/898586/topics/topic04.pdf

So this is what we got (if we say our newton fluid stays laminar) :

F = η * A * v / l ..with the dynamic viscosity η with the unit Newton-sconds-per-square-meter [ (N * s) / m² ]= [Pa * s] or Pascal seconds. Water at 20°C has a viscosity of 1 mPa·s (milli-pascal seconds), engine oil has 100 mPa s at the same temperature.

we wanna know drag torque, so with the radius r =0.04m we make it Tdrag = r * η * A * v / l

So for water between stator and rotor we get:
T = 0.04 m * 0.001 Pas * 0.0125 m² * 8.33 m/s / 0.0005 m = 0.0083 Nm of drag...this seems acceptable.

With engine oil we get 100 times the drag, which would be 0.833 Nm which is not acceptable IMO


The rotor doesn't get much heat sinking ATM, but it get's heated by (small but existent) eddy currents. Cooling it as well is no drawback in my eyes.
 
According to this page http://www.substech.com/dokuwiki/doku.php?id=mineral_transformer_oil
transformer fluid has 22mPas at 20°C and 2.6 mPas at 100°C. Sound good to me

Edit:
this is 0.1823 Nm or drag at 20°C and 2000rpm. This is additional loss of 38.2W ... So the no load consumption may raise by double of that, say 80W

At 100°C there is only 1/8 th of this loss..
 
Polyesterimide (german word, i dont know the english term) are a base material for the varnish used for magnet wire. Which again is based on ethylene glycol (i read on wikipedia)

So glycol seems to dissolve our wire varnish. Not a good idea at all :lol:

I tend more to mineral isolation oil /transformer oil. It's designed for exactly our purpose (almost) Sadly it comes in 20L canisters :shock:

Eurolub Transformer oil (10 mm²/s at 40°C), this is ~ 8.7 mPas
datasheet http://www.eurolub.com/transformatorenol.html?___store=en&___from_store=de
shop: http://profiteile.de/eurolub-transformatorenoel-20-l

Shell Diala S2 (11.5mm²/s at 40°C), this is ~ 10mPas
datasheet http://s05.static-shell.com/content/dam/shell/static/ind/downloads/lubes-b2b/diala/shell-diala-s2-zui.pdf
shop: http://www.cs-schmierstoffe.de/shell-diala-20ltr/p-31.html
 
You must log in or register to reply here.

Similar threads

Hugh-Jassman
How to read cryptic motor graph?
Replies
5
Views
1,314
Hugh-Jassman
Hugh-Jassman
G
My First Build : MILDA MK 1 /out of Commision Please Help!
2
Replies
27
Views
3,291
glenn0010
G
G
Custom BLDC Controller Closed Loop
2 3 4
Replies
77
Views
9,666
glenn0010
G
bigbore
Looking for a brushless powerful motor for my e-scooter
Replies
16
Views
4,605
bigbore
bigbore
B
Big hub motor 308mm 56 poles
Replies
24
Views
5,579
made_in_the_alps_legacy
made_in_the_alps_legacy
Back
Top