ESE's works engine tuner

[F5 Dave]

That sounds great, but why would you reverse the flow? As you said but cold in the rear low. Been covered a few times. Std system is like squirting a hose in a bucket

[Frits Overmars]

With no attempt at cancelling the crank oscillating forces internally, then the cases/bearings are being severely hammered even if it cant be felt by the rider due to some clever isolation techniques.

Can you imagine the internal hammering in a Rotax tandem-twin? The two contrarotating crankshafts doubled as balance shafts for each other, so their balance factor was almost 100 %. I had gathered some good experiences with 125 cc Rotax singles and when the 250 cc tandemtwins arrived, I expected to get them running without many problems. The first problem was cracking disc valves that had never caused any problems in the single...

wob on the yami twins would I gain anything worth while if i reverse the water flow and take it from the rad to the top of the head at the front of the cylinder then weld on some spigots as low as I can at the cylinder rear just above the reed blocks ? this is how some of the older ktm water flowed....
it shouldn't be much trouble to do up a impeller with reverse fins...

Peewee, you might try to figure out why KTM bade farewell to their old layout. Meanwhile I'll try to figure out what you hope to achieve by reversing the impeller fins. The impeller efficiency will become hopeless unless you also reverse its direction of rotation.
In itself that is quite simple: you can push start your Yamaha backwards. But riding it with zero forward and six backward gears will be a challenge.
And the pump will still function less than optimal because its spiral casing isn't suited for the reverse rotation.

[wobbly]

The Yamaha LC/TZ water routing is a complete disaster,but reversing it wont solve anything.
I made CNC impellers 50% wider, and simply welded a spacer plate on the pump cover to increase the flow rate.
And the only way to get the water to flow correctly is to weld a flat 6061 plate on the top deck.
This needs really good welding around the bore, with the correct rods to suit the plating.
Then a single inlet down low at the rear brings in cold water, that is forced to flow over the transfers ,and then around the exhaust.
I put 4 holes in the plate equivalent to the exit pipe area, at the front of the cylinder, 1 each side of the 2 exhaust ducts.
This then forces water up into the front of the head, where it then flows around the inserts, and out at the rear of the head.
The thermostat housing was cut off and a plate welded over as there is a proper bypass setup near the pump entry..
This system works perfectly and just needs a small ( 3mm ) hole at the highest point in the middle of the plate to allow air to escape from
under this cylinder high point up into the head that has the exit - so its all self bleeding.

[adegnes]

Ages ago (I think but could be wrong)either 2Tinsitute or Ken Seeber suggested a cheaper alternative for small quanities of a tunsten alloy.
it was on Ebay for these things
http://www.abc-pinewood-derby.com/weights.php
http://pinecar.woodlandscenics.com/s...m/P3915/page/1
http://www.maximum-velocity.com/tungstencylinders.htm
http://www.ebay.com/itm/Incremental-...-/322004729139

You could also check out your local angling/fishing store.

I've pulled the crank.
Playing with nuts and washers taped to the counterweight tells me I need about 24-25g of added weight where I plan to drill the holes(two in each web close together opposite from the crank pin).

Both crank webs are 14mm thick.
A common size for the Derby Tungsten slugs are ø9.5mm x 11mm

Two ø10mm x 14mm holes in each web = about 34.5g removed
four ø9.5mm x 11mm tungsten slugs = about 60g added

60 - 34.5 = 25.5

small adjustments to the slugs with the grinder and we should be there.

Is gluing them in with epoxy certain death you think? The slugs seems to be cast, and are probably not super round.

[wobbly]

Forget gluing - have a light press fit and you can melt the web and tungsten together with tig, two small arcs of
weld with filler rod each side works perfectly.

[peewee]

frits thnx for bringing me back to reality. I forgot the reverse fins would be useless without the shaft turning in reverse also . wob I think ill just keep the water flow path as it is now. maybe not ideal but if I get to involved with trying to re-engineer every aspect of these dinosaur engines ill be a old man before its up and running

[wobbly]

Problem is that using the stock parts and water path its going to be impossible to keep the temp down where it should be.
They run hot as hell, and loose huge amounts of power in the process.

[TZ350]

What you are after is the average temp of the pipe wall material. You would need to measure the hottest ( at the header ) then the coldest ( in the mid ) - then it gets hotter again as you travel down the rear cone. Problem is that there is far more area of cold surface in the middle, so sadly some fudge factor guesstimating is needed.

Our method was not very scientific, just a bit of curious fun, but it looks like with a bit of effort and following Wob's suggestions one could easily make a good temperature map along the pipe.



Pipe temperature of a Honda RS125 taken during a dyno pull from 7K rpm up to peak power and then run back down again.

[jfn2]

TZ350:
The way I see it, and please tell me if I'm wrong and how I'm wrong, either the pipe in your video is not matched up with your RS125 engine or the engine is not tuned correctly because the temp at the header (which should be the hottest temp area) is only 337 degrees at peak and not around +/- 500 degrees. I don't know where the bottom of your power band is but I don't think it's anywhere near the 325 degree mark. If I am right how would you go about correcting this setup? Also if you were to check the header temp with either a heat gun or a heat surface temp sensor where would you place the sensor? My guess is about 23cm (9") from piston face. The center sensor would be placed in the center of the belly (dwell) section and the last sensor at the center of the last cone ( center of the complete cone). Any one have any thoughts on this? Thank you.

[wobbly]

The max temp of 337*C is about 1/2 the way down the diffuser - nowhere near the header.

[jfn2]

wobbly:
Sorry my bad. So is this a good reading for this area?

[TZ350]

The RS was making 41 DynoJet rwhp, so about average. It was just interesting to see how the temperature changed with rpm.

[wobbly]

The temps we use in EngMod are the reflected wall temps the exhaust gas sees on the INSIDE as it travels down the pipe.
A heat gun shows whats going on on the outside,with plenty of slowly reacting insulation in the form of oily crap on the inside
and rusty crap on the outside surfaces.
As I said you have to guestimate whats going on from a series of heat gun readings ( this gave me the 100* split number ), and then use an actual dyno result to work backwards
in the sim to get a realistic curve shape in the right rpm band.
This is how I ended up with the 325/425* result that is close ballpark for many high performance race engines.

[ken seeber]

Sometime ago Hooser got me excited. Nah, nothing to do the (or his?) whole chicken thing, but labyrinth seals (see pic). HP just waiting to be unleashed?
So, got a Ø20*Ø35*7 teflon lipped seal and wrapped it with a paper strip, slid it over a crank and weighted it up so it was enough to overcome static friction and slowly rotate around the shaft with a 90 gm weight. (see pic).
Working backwards, this level of friction equates @ 14k rpm, to equal 0 .0226 kW. Stuff all really.
However, this is without and pressure acting on the seal. I guess the next step would be to mount 2 seals in a sleeve with varying applied pressures from within a drilling in the shaft. See pic. I guess one could do this based on crank angle vs pressure and do some sort of averaged calc of the frictional power over the full 360 deg cycle

[Hemi Makutu]

Aye, tip some bloody nitro-methane in eh fella, its been research proven.

http://firebrand.me.berkeley.edu/thesis/chen_thesis.pdf