Thinking about a Bus/RV

[asicer]

Ah. All the old Alfas come to mind.

Getting OT here but there just enough details in this pic to pique my interest, so...

Which old Alfa are we looking at?

 

[ElPaso Pilot]

Getting OT here but there just enough details in this pic to pique my interest, so...

Which old Alfa are we looking at?

That one was a random Google search:

1964 Alfa Romeo Giulia Sprint Gt

https://www.tommycarclassic.com/en/vehicle/alfa-romeo-giulia-sprint-gt-1964/#prettyPhoto[12499]/2/

but as I recall, finned pans seemed to be an Alfa thing. Had a friend with a Spider that looked similar.

Always worried as those fins were low, exposed, and poorly protected.

 

[Ted]

Always worried as those fins were low, exposed, and poorly protected.

The BMW R1150GS air-head engine (i.e. oil cooled) that my friend is letting me borrow for a while has a finned oil pan. Also fairly low, and for an engine intended to go off-road, seems like it could be a source of easy damage. Of course those engines also have spark plugs on the bottom of the heads (as well as side) which seems to be a source of complaint by adventure riders who like to ford rivers and streams with their bikes. However, as hard of a life as this particular motorcycle has had, it doesn't seem like there are any missing fins. I think this is its second engine, though.

 

[Spring Ford]

Always worried as those fins were low, exposed, and poorly protected.

I would imagine that the fins stiffen and protect the sump pan.

Here is the humble Mini engine. Fins go all the way along the bottom.

 

[Ted]

Today I started going through the cooling system on the bus. While I knew I was going to do it over the winter, I'm glad to be working on it now. One hose was leaking, and the hoses appear to be almost all original. Fortunately the one hose that is definitely not original is also the one that I can't get to, which connects the water pump to the engine. Incidentally, I have no idea why Cat put the water pump where they did - on what otherwise is a very nice engine design, the water pump location and interface isn't very good.

One of the hoses from Freightliner is the wrong hose, and another hose that they told me to get 3/4" hose for (generic heater hose) isn't 3/4", and is clearly a shaped hose. So, I'm going to go back to them and see if they can fix those. But otherwise I can clean things up and start putting them back together. I also have to see which clamps I want to reuse vs. replace. For any of the hoses 3/4" diameter or greater, they actually used very high quality hose clamps that worked just fine. The smaller hoses (and clamps) I'll definitely be replacing, though.

The transmission cooling I've figured out what I'm going to do, and it should be a good first step for my eventual goal of converting the bus over to electric cooling fans entirely over the winter. I've got one (pretty massive) Derale transmission cooler with built-in fan that's designed for remote mounting (i.e. mount it anywhere). It's intended for use in areas where direct airflow may not be feasible, which is good since there's really not a good place for me to add a cooler with decent airflow on the bus. I've got a second one coming, the two will run in parallel and the transmission fluid will go through them, with a thermostatic switch to control the cooling fans. I'll then bypass the stock transmission cooler entirely. This does two things:

1) Eliminates the failure mode of the cooler failing internally, allowing transmission fluid and coolant to cross-contaminate.
2) Provides all of the cooling capacity that was being used on the transmission to cool the oil and engine

Since the transmission got first crack at fresh coolant from the radiator, when the engine would get warm it would pretty much always be because of the transmission. And these transmission generate quite a bit of heat.

This will require new transmission lines to be made up (the old ones are very clearly original, and something I wanted to replace anyway), plus some adapters. So I'll get all of those figured out. And then I'll carry the old ones with me, since the old cooler will still be in situ. That way if there's a problem with the new setup, I can revert back to old.

Most of the posts you see recommend doing multiple flushes of the coolant when you change it - run water through it first to get any junk out. This coolant really didn't look bad. It was starting to get a bit cloudy/rusty, but it was still the appropriate red color and I got about 7 or 8 gallons out (for what's supposedly a 10ish gallon system). So, I figure that over the winter when I do some additional work like the electric fans and likely changing the coolant hoses going all the way to the front of the bus, I'll probably be draining some of the coolant again, and that will get me close enough to new. Supposedly it was replaced 6 or 7 years ago (I forget the exact time) and Cat recommends 6 years, so this isn't too far off.

I'm excited for this trip just to see how the cooling system does!

 

[tmyers]

What Temps do normally see on your transmission, Ted

Sent from my SM-N970U using Tapatalk

 

[Ted]

What Temps do normally see on your transmission, Ted

So keep in mind our bus is about 28k lbs with the Land Rover behind it and stuff in it. On this last trip, we were seeing consistent temps in the 195-210F range. While allowable per Allison, ideal is more in the 175-185 range. I start to notice it shift a little funny at the higher temps, so that tells me the trans doesn’t like it.

After I get this installed I’ll do a video on what I did and why with some diagrams.

 

[NealRomeoGolf]

I have no idea why Cat put the water pump where they did

I am no engineer, but since these are from CAT's on-highway line, I wonder if it has to do with it was there best for 18 wheelers in the cab and not so great for RVs in the back. I dunno.

 

[Ted]

I am no engineer, but since these are from CAT's on-highway line, I wonder if it has to do with it was there best for 18 wheelers in the cab and not so great for RVs in the back. I dunno.

That may have been it. Yeah to be fair to the design, it’s very serviceable on the whole, which is one of the things I like about the engine.

it’s just an interesting design decision when other items (like the air pump and oil cooler) are integrated into the engine, no belt no nothing.

I can’t wait to see how she performs after I free up some of the horsepower with the electric fan conversion!

 

[Ted]

Today’s progress: adding the air scoop for the turbo. We’ll see if this works as expected or not.

 

[Ted]

Later this week I'll have one other fun project, bonding heat sinks to the oil pan. I'm looking at four different adhesives for this to use:

1) JB Weld (standard)
2) JB Weld Marine Weld
3) Permatex Cold Weld
4) Permatex Steel Weld

The trick with this is that I'm looking at bonding stamped steel (the oil pan) and aluminum (the heat sinks) which can be a bit tricky. The respective companies say that options 2 and 3 are the best for working with aluminum and in an auto environment with exposure to oils, etc. They have a lower bond strength (something like 3k psi instead of 4-5k) but the reality is this is not a high strength application, more important is that it can withstand the vibrations and potential exposure to fluids.

I'm not a specialist when it comes to these adhesives, curious if anyone has a suggestion. I'm likely going to try 2 and/or 3 just because they seem to be the best fits on paper.

 

[Bill]

but the reality is this is not a high strength application, more important is that it can withstand the vibrations and potential exposure to fluids.

And TCE mismatch as the materials expand at different rates. A strong yet compliant bonding material is the right call here.

 

[Ted]

And TCE mismatch as the materials expand at different rates. A strong yet compliant bonding material is the right call here.

Yes, those different rates of thermal expansion are going to be a big one as well. I think that remaining compliant is one of the benefits listed for those preferred (2 and 3) options as well.

 

[jsstevens]

Yes, those different rates of thermal expansion are going to be a big one as well. I think that remaining compliant is one of the benefits listed for those preferred (2 and 3) options as well.

Does the adhesive conduct heat? If not, this may not let the heat sink do it’s job unless there’s lots of metal to metal contact. In electronics we use heat sink compound but it’s a grease not an adhesive.

 

[Bill]

Does the adhesive conduct heat? If not, this may not let the heat sink do it’s job unless there’s lots of metal to metal contact. In electronics we use heat sink compound but it’s a grease not an adhesive.

Excellent point!

 

[Ted]

Does the adhesive conduct heat? If not, this may not let the heat sink do it’s job unless there’s lots of metal to metal contact. In electronics we use heat sink compound but it’s a grease not an adhesive.

I've been looking into that as well, and it doesn't seem to be published much for any of them. While the thermal conductivity of any of those isn't going to be as good as a heat sink compound, none of them seem to be insulators, and they'll be fairly thin layers. So basically my takeaway is they won't hurt things.

 

[jsstevens]

I've been looking into that as well, and it doesn't seem to be published much for any of them. While the thermal conductivity of any of those isn't going to be as good as a heat sink compound, none of them seem to be insulators, and they'll be fairly thin layers. So basically my takeaway is they won't hurt things.

I would get as much metal to metal contact as you can then. If the adhesive doesn’t conduct heat then it’s an insulator (albeit a thin one) and you’ve lost the benefit of the sink. What does the manufacturer recommend for installation?

 

[Ted]

I would get as much metal to metal contact as you can then. If the adhesive doesn’t conduct heat then it’s an insulator (albeit a thin one) and you’ve lost the benefit of the sink. What does the manufacturer recommend for installation?

That I don't know. Another idea I've considered is just doing the adhesive around the outer perimeter of the sinks, and then putting some sort of thermal conductive paste/etc. in the middle. Basically have one do the bonding and the other do the conducting.

 

[jsstevens]

That I don't know. Another idea I've considered is just doing the adhesive around the outer perimeter of the sinks, and then putting some sort of thermal conductive paste/etc. in the middle. Basically have one do the bonding and the other do the conducting.

That sounds like a decent approach. I suppose you could drill and tap the pan, but I wouldn’t like extra leak potential.

 

[Ted]

That sounds like a decent approach. I suppose you could drill and tap the pan, but I wouldn’t like extra leak potential.

The pan is just stamped steel. It's fairly thick as far as stamped steel pans go, but it definitely couldn't be tapped. If I were to do something like that, it would get more complicated fairly quickly.

So maybe I'll do that - order some heat sink paste and try the two combined, dab in the middle for the actual conductivity and then the adhesive around the outside to hold it on.

 

[Ted]

So @jsstevens (and anyone else who knows more about computer stuff than me) - when you put on thermal paste/thermal adhesive, how do you apply? I'm assuming just an appropriately sized dab in the middle of the heat sink? One of the things I'm thinking about is trying to make sure that it wouldn't interfere with the actual adhesive.

 

[jsstevens]

So @jsstevens (and anyone else who knows more about computer stuff than me) - when you put on thermal paste/thermal adhesive, how do you apply? I'm assuming just an appropriately sized dab in the middle of the heat sink? One of the things I'm thinking about is trying to make sure that it wouldn't interfere with the actual adhesive.

You try to squeeze it out to cover the contact area. You want the most contact area covered to transfer the heat.

and you definitely don’t want it in/under the adhesive as it’s like grease and will interfere with adhesion.

 

[Ted]

You try to squeeze it out to cover the contact area. You want the most contact area covered to transfer the heat.

and you definitely don’t want it in/under the adhesive as it’s like grease and will interfere with adhesion.

Yeah, so I'm thinking some balance in the middle is probably what I want to do. Basically, do an amount in the center to make sure that it has the conductivity, but not so much that it would interfere with the adhesive.

I suppose the other option would be to reverse it - do a good dab of the adhesive in the middle, and then get the outer edges with the thermal paste. I tend to think that would be more likely to have road stuff and water eventually wash that paste away, though - with the adhesive on the outside it's sort of holding the paste inside.

Looks like Artic MX-4/5 is considered the best?

 

[Ted]

I guess another question is what's considered "good" thermal conductivity? This Artic Silver stuff claims something in the 6 W/m-k (seen different claims on the internet). I'm seeing some that claim that, but most claim something in the 1.2-1.5 range. This also looks interesting:

https://www.amazon.com/AVNTKER-Adhe...626098902&sprefix=thermal+adh,aps,241&sr=8-13

Claims a 1.5 W/m-k thermal conductivity, has some adhesive properties. Could use that and then add some of the other adhesive around the outer edges, more like a weld style.

 

[masloki]

So @jsstevens (and anyone else who knows more about computer stuff than me) - when you put on thermal paste/thermal adhesive, how do you apply? I'm assuming just an appropriately sized dab in the middle of the heat sink? One of the things I'm thinking about is trying to make sure that it wouldn't interfere with the actual adhesive.

In the PC world you have 2 surfaces nearly perfectly milled. You apply a dab, and scrape with a credit card. That way it is only filling the irregularities with ultra thin layer. I think you are on the right track with adhesive thermal compound on the center, and regular adhesive on the outside. I assume if you lose the heat sink outright you are impaired, not dead in the road. I’ve seen thermal tape which may fit into your solution too.

 

[Ted]

In the PC world you have 2 surfaces nearly perfectly milled. You apply a dab, and scrape with a credit card. That way it is only filling the irregularities with ultra thin layer. I think you are on the right track with adhesive thermal compound on the center, and regular adhesive on the outside. I assume if you lose the heat sink outright you are impaired, not dead in the road. I’ve seen thermal tape which may fit into your solution too.

Any suggestions on the thermal tape (I linked one above) or how well they work?

To answer the second part of your post, if the heat sinks fall off, I am no worse off than I am today. The oil pan has absolutely no design features that make it try to shed heat. The goal is for this to add cooling capacity to some extent.

 

[masloki]

Any suggestions on the thermal tape (I linked one above) or how well they work?

To answer the second part of your post, if the heat sinks fall off, I am no worse off than I am today. The oil pan has absolutely no design features that make it try to shed heat. The goal is for this to add cooling capacity to some extent.

Oops, didn't see it was a tape. 3M has 4 to choose from, where the thickness varies, and I'd put a lot more trust in the Minnesota guys than the one you linked. My notes are based on PC building, and I assumed paste was better than tape for any application where thermal conductivity was key. The 3M stuff has pretty good stats though.

 

[Ted]

Oops, didn't see it was a tape. 3M has 4 to choose from, where the thickness varies, and I'd put a lot more trust in the Minnesota guys than the one you linked. My notes are based on PC building, and I assumed paste was better than tape for any application where thermal conductivity was key. The 3M stuff has pretty good stats though.

I'd definitely go with 3M over much of anyone else. It seems their tapes have a rating of 1.5 W/m-k. So that gets me back to my previous question - what is considered a "good" number? Especially looking at bonding steel to aluminum.

 

[jsstevens]

I liked the tape option for your application. Maybe supplement with JB Weld around the edges

 

[Ted]

So it looks like the thermal tapes say 0.2mm thick (i.e. really thin). That makes me wonder if it would get a good enough bond to actually do much heat transfer. The past would have the advantage of filling the voids better. Thoughts there? This oil pan is not a super smooth surface, even with prep.

 

[Spring Ford]

Google [thermally conductive adhesive] gets good looking hits.

 

[Spring Ford]

It seems their tapes have a rating of 1.5 W/m-k. So that gets me back to my previous question - what is considered a "good" number?

"The thermal conductivity of steel is measured at approximately 45 W/(mK), which is extremely low compared to copper and aluminum that exhibit a thermal conductivity value of 398 W/(mK) and 235 W/(mK) respectively"

I guess this is analogous to electrical conductivity?
Using your example - 1.5W/(mK)

A 1 meter cube of the material will conduct 1.5W per Kelvin.

Think of the cube as having two 'dimensions' a length and an area.

As the length reduces the heat passed will increase proportionally
As the area reduces the heat passed will decrease proportionally.

So you have about say a square foot, or 0.1 sq m.

With tape 1/5000 of a meter thick you would get

1.5 * 5000 /10 = 750 watts per degree.

With a 60K difference across it - (100 degree day) it might pass

45,000 W - if the fins dissipated all of the offered heat and held themselves at the air temp.

What does that mean? No idea

20% of the engine power (200kW? - 275HP).

I am far from sure if the above is anywhere near right.

Edit - 0.2mm is 1/5,000 of a metre not 1/500 - fixed.

 

[Ted]

What does that mean? No idea

20% of the engine power (200kW? - 275HP).

I am far from sure if the above is anywhere near right.

I think this basically sums up what I've figured from looking at the units and calculations, especially the "No idea" portion.

The engine is rated at 330 HP at the crank flange, which means it makes more thermodynamic power than that at the combustion chambers and some portion of it goes into the exhaust, cooling system, and oil in the form of heat. Plus the amount that goes into rotating friction and driving the oil pump.

Let's say that I'm running it at 65% power going down the road (which isn't too far off for a highway run), so roughly 220 HP or some number less than that kW.

Empirically, looking at air-cooled engines (which of course have a significantly higher delta T at the cylinder) I would tend to believe the level of cooling/heat sink I'm looking at is going to be sufficient for my cooling purposes. Even looking at the BMW R1150RT that I'm borrowing currently, the oil pan has cooling fins cast/machined in to the oil pan, so that must do enough to cool the oil that it's worth them spending the money on. The real goal then is to optimize the heat transfer between the oil pan and the aluminum heat sinks. There will be a good amount of turbulence with the oil inside the oil pan while running, so that will take care of heat transfer at the oil pan surface. Then I just make sure the heat sinks are bonded and as thermally conductive as I can.

I ended up deciding I'd use some of that Artic MX-5 thermal paste since it seemed to have the best ratings and overall considered very good. Then I'll put some sort of 2-part adhesive around the outer edges of the heat sinks, put them on, and we'll see how well they hold up and if I can notice a difference.

We'll see what happens. This will be a fun experiment.

 

[Bill]

Even looking at the BMW R1150RT that I'm borrowing currently, the oil pan has cooling fins cast/machined in to the oil pan, so that must do enough to cool the oil that it's worth them spending the money on.

I'm inclined to believe the finning works very well, and here's why. Riding at highway speed on cold days (temps in the 30's), the oil temp gauge would still run mid-gauge as the vernatherm in the oil cooler regulated the flow to keep the engine temp centered. Ride on a cold rainy day, however, where the front tire was spraying cold water up onto the oil pan, and the bike would only manage to keep one bar on the temp gauge, well lower than normal. That tells me the vernatherm was shut fully off and the cold rain/wind on the oil pan was keeping temps down. Very repeatable, the cold running would happen anytime it was raining and below 45* or so.

 

[Ted]

I'm inclined to believe the finning works very well, and here's why. Riding at highway speed on cold days (temps in the 30's), the oil temp gauge would still run mid-gauge as the vernatherm in the oil cooler regulated the flow to keep the engine temp centered. Ride on a cold rainy day, however, where the front tire was spraying cold water up onto the oil pan, and the bike would only manage to keep one bar on the temp gauge, well lower than normal. That tells me the vernatherm was shut fully off and the cold rain/wind on the oil pan was keeping temps down. Very repeatable, the cold running would happen anytime it was raining and below 45* or so.

That's a good empirical analysis, Bill. We'll find out once I get it on the road.

Once Amazon delivers my second transmission cooler, I can work on getting that whole system together for the trip as well. If I don't screw the whole thing up, the system should be running much cooler.

 

[ElPaso Pilot]

I'm inclined to believe the finning works very well, and here's why. Riding at highway speed on cold days (temps in the 30's), the oil temp gauge would still run mid-gauge as the vernatherm in the oil cooler regulated the flow to keep the engine temp centered. Ride on a cold rainy day, however, where the front tire was spraying cold water up onto the oil pan, and the bike would only manage to keep one bar on the temp gauge, well lower than normal. That tells me the vernatherm was shut fully off and the cold rain/wind on the oil pan was keeping temps down. Very repeatable, the cold running would happen anytime it was raining and below 45* or so.

So, my takeaway here is that Ted needs to build his finned oil pan with a water spray setup, for a cooling boost on those long hot hill climbs.

He always seems game to lower himself down into another experimental rabbit hole!

 

[Bill]

So, my takeaway here is that Ted needs to build his finned oil pan with a water spray setup, for a cooling boost on those long hot hill climbs.

He always seems game to lower himself down into another experimental rabbit hole!

And hey, why not water injection into the engine going up hills as well? He can roll coal like an old B-52 on takeoff.

 

[Ted]

So, my takeaway here is that Ted needs to build his finned oil pan with a water spray setup, for a cooling boost on those long hot hill climbs.

He always seems game to lower himself down into another experimental rabbit hole!

And hey, why not water injection into the engine going up hills as well? He can roll coal like an old B-52 on takeoff.

I had considered building a finned oil pan custom for the bus, of course the realities of that are far too much time that I don't feel like doing, hence the heat sink idea. The water cooling on the oil pan I didn't think of (and wouldn't do), but I had considered water injection/water cooling on the intercooler to help lower induction air temps. I ultimately decided I didn't want to have to refill the bottle or take water away from the household water supply, which we are usually trying to preserve.

 

[Ted]

I'm starting to think about my next crazy idea, which is how I want to control the cooling fans when I convert the bus to electric fans over the winter (assuming this drive goes well).

I'm looking at 5-6 12" diameter, 2000 cfm electric cooling fans, probably in a pusher configuration ahead of the intercooler (the intercooler goes in before the radiator). So, say 10-12k CFM of air, which should be sufficient to cool the thing under all conditions. I have the electrical capacity for this. But I don't want that many amps (especially that many amps of inrush current) coming on all at once, and it would be silly to do so as that much cooling isn't what the bus will always need, or even often need. I also may find that I want the fans to be turned on under certain conditions to keep intercooler temps below a certain number. Not that I'm hugely concerned about induction air temps on the whole, but right now the intercooler receives constant airflow of the fan, and that would be changing.

This ends up resulting in some semi-complex level of control. Prevost does this today, as do some other bus makers:

In thinking about the control, I think a Microsquirt like what I have on the Cobra might be overkill, but it might also be a good way to work things. I don't need to run an engine with it, but it has programmable discrete outputs, and could be set up to work with a coolant temp sensor (extra ports exist on the Cat's thermostat housing I could use), induction air temp sensor, even a boost sensor (or some other things). Then I could stage the fans turning on and off at different temperatures so that say at 195F coolant temp I turned on 2 fans, 198F 2 more fans, and 202F all 6. I could then do some "or" gates in there to allow fans to turn on for induction air temperatures, things like that.

A Microsquirt with harness is in the $350-400 range, and I feel like there should be a cheaper solution I could use. But it might make sense and the programmable capability (plus familiarity) might make it make sense.

Anyone know of another option? I was thinking from what I understand a Raspberry Pi might have the capability, but I really know nothing about the things and am not a software programmer myself.

 

[SoonerAviator]

So, my takeaway here is that Ted needs to build his finned oil pan with a water spray setup, for a cooling boost on those long hot hill climbs.

He always seems game to lower himself down into another experimental rabbit hole!

Why stop there? I say he needs a whole new custom oil pan with cast-in fins and an internal water jacketing for a cryo-cooling setup, lol.

Anyone know of another option? I was thinking from what I understand a Raspberry Pi might have the capability, but I really know nothing about the things and am not a software programmer myself.

I'm no Raspberry Pi programmer, but I'd think that would be the perfect application for one. One or two temperature inputs (coolant temp/intake air charge temp) with some relatively simple code to determine which fans come on under what conditions. No need for something as extravagant as MS to run a fan array.