112.8mph Lucky Poos Needed

RiftAmps

So, what causes a spun bearing?

I've been doing a lot of research into this because I want to learn what's going on, what happened, and how we can prevent it.

Engine Builder Magazine says:

A spun bearing is usually the end result of a chain of unfortunate events. These include high operating loads, excessive heat and a loss of lubrication.

Spun bearings are often blamed on a loss of oil pressure, but there can be numerous contributing factors that eventually cause the bearing to seize and spin. Normally, bearings allow the crank glide on a hydrodynamic film of oil between the bearing surface and journal. As long as there is a steady supply of clean, fresh oil flowing into the gap between the bearing and journal, everything is fine. The film only has to be a few microns thick to maintain protection. But if there is a sudden loss of oil pressure and flow, especially at high RPM, the protective film can disappear in an instant. Once that happens, you get metal-to-metal contact between the bearing and journal surfaces, which causes scuffing, increased friction and heat.

So, two questions to answer;

1. Did we have enough oil in the engine?
2. Did we miss anything else?

The answer to #1 is easy - Yes, we had EXACTLY 3.8 litres of oil in the engine. I know this because we had carefully measured out the right amount of mineral oil and put that in.

#2 is a little more embarrassing to answer, but only because it points to user error rather than something else.

I've spoken to my local engine machinist, a dedicated MX5 engine builder, and the owner of an MX5 race team. They've all told me what they think happened, and I'm afraid it's all my fault.

You see, when you're running in an engine you're using either mineral oil or a dedicated break-in oil, designed to help the piston rings seat to the cylinder walls. These oils are great for that very specific job, but don't offer a lot of protection at high-rpm because they simply lack the film-strength at those loads. It's why Blink Motorsport said we shouldn't use more than 6000rpm on track and why Basset Down Balancing won't do any power runs on their dyno if the engine is filled with mineral/break-in oil.

So, we had our brand new engine filled with 15w-40 mineral oil and ran it all morning up to our 7250rpm rev limit. It was always going to fail. Lesson learned.

How do we prevent it from happening again? Fortunately, lowering the rev limit on the car is quite easy with our ME221 ECU - just duplicate our race map, enter a lower rpm figure for the limiter and load that onto the car. Once the engine has done a full day of running we can change the oil to fully-synthetic and reload our full race map.

crunchman

Expensive lesson!

strtdv

RiftAmps said:

strtdv said:

Also, not sure if you intended to imply the coolant leak caused that big end bearing failure?

Not at all, just the order in which I updated the thread.

We spun the bearing at Silverstone in Feb, removed the engine from the car and swapped the new cylinder head onto the old block and then reinstalled that into the car. It was then that we realised that we must have damaged the drain plug.

Ah, that makes much more sense!
And I see you subsequently got to the bottom of it, an expensive lesson indeed! As you say though, easy enough to set a software limiter on a dedicated "running in" map.
I'm looking at a remap for road purposes to get a bit more low end shove (for a big turbo engine the power is quite focused up at the top end over 5900rpm when the limiter kicks in at 6750), and I see a lot of them add a progressive limiter that lifts as your oil temp gets up

RiftAmps

strtdv said:

Ah, that makes much more sense!
And I see you subsequently got to the bottom of it, an expensive lesson indeed! As you say though, easy enough to set a software limiter on a dedicated "running in" map.
I'm looking at a remap for road purposes to get a bit more low end shove (for a big turbo engine the power is quite focused up at the top end over 5900rpm when the limiter kicks in at 6750), and I see a lot of them add a progressive limiter that lifts as your oil temp gets up

The MEITE software allows you to do the same thing, although it uses the coolant temp as MX5s don't have an Oil Temp sensor.

Here are the stock settings for a road car, the limiter is set to 6500rpm until temp is above 20c then ramps up to 7400rpm at >40c. Safety limiter set at 4500rpm if the temps go over 100c.

The Soft Cut RPM Offset is how many revs below the hard cut you want the ECU to start progressively cutting a percentage of fuel/spark from 0-100% as you approach the hard cut. Stock is 200rpm and adjusts only the fuelling. You can choose, spark, injection, or both.

Below are the limiter settings for our race car, a simple 7250rpm at any temperature. This allows the car to continue revving as intended if the temp sensor fails and sends an erroneous signal to the ECU.

The soft cut is set to just 50rpm below the hard cut.

And finally the new 'running in' settings set to 6050rpm. To change the values, you simply just enter them into the table on the left and press enter.

RiftAmps

Crankshaft

Bad news - When I was stripping the block down to remove the crankshaft, I noticed some odd wear patterns on the main bearings. You can clearly see not just light scoring, but also spotting where the upper layer has worn through. The picture below shows three of them.

I took the crankshaft over to the machine shop for polishing but when we placed it onto the machine, we discovered that it was bent like a banana - not visual to the naked eye but in crankshaft terms, it was borked.

The spec sheet says that a maximum of just 0.002" is allowed and there are specialist machine shops that might be able to straighten them but ours was a massive 10x the maximum amount measuring in at 0.020" (20thou). The high frictional heat and rotational loads at 7250rpm is enough to bend a crank - who knew?!

As always, my tactic of buying up all the used spares that I can has once again paid dividends as I had another crank at the workshop. We checked that just in case and all was OK so it's now been polished and is ready to be installed.

New bearings have been ordered so I'll be able to start the rebuild soon.

sweepy

At least you have plenty on to keep you out of mischief 

RiftAmps

Rebuild

New bearings, crankshaft, and a replacement conrod for cylinder #3 and it's back together fitted with our spare cylinder head that's now the same spec as the other (Piper 285H camshafts, 3-angle cuts etc). The build process was the same as last time so didn't take any more photos apart from the finished engine below.

At the moment it doesn't look like we'll be racing until August but will try and get a track day in beforehand once Motorsport UK start to issue event licenses again.

RiftAmps

Coolant Reroute

A popular mod for the MX5 is to revert the cooling layout to that of the original design, found in the Mazda 323. In that car, the cold coolant enters the engine at the front, through the water pump and exits at the rear from the cylinder head. When Mazda turned this engine longitudinally and put it into the MX5, they moved to the coolant exit to the front, leaving the rear of the engine to run hotter as it is not properly cooled by coolant flow. You would think this would be a problem but in reality, it's really not and most MX5s have run like this forever, without failure.

When people start to mod their cars, however, especially with forced induction setups, cooling becomes an issue and thankfully a few aftermarket companies now offer kits to reroute the cooling as intended.

Original layout

Reroute

Since installing our aluminium radiator we've not suffered any cooling issues, but we have created ourselves a little issue that has bugged me. You see, we've removed the heater core from the car and simply run an extra coolant hose to connect the send and return pipes together. It's not pretty but because of the current hardware, there has been no real way to neatly 'delete' the heater core and install the reroute.

RiftAmps

Here are the two heater core pipes, the 'send' is on the right and the 'return' is on the left. Two short hoses connect them to the heater core through the firewall.

The Return pipe hangs off one of the exhaust manifold studs and connects to the Water Pump inlet pipe.

The HC return pipe presses into the inlet pipe and is held in with nothing more than an o-ring. Because of this design, you must have the HC return pipe installed otherwise coolant would piss out everywhere.

RiftAmps

A secondary issue is that the Water Pump inlet pipe contains an 8mm off-shoot pipe that connects to the thermostat housing, and onto the idle-speed air valve on the inlet manifold, via another 8mm pipe and hose. We need to retain these.

RiftAmps

Thankfully, a small British firm called G19 Engineering makes a water pump inlet pipe from billet aluminium that is designed for those who want to delete the heater core. Crucially, it also has an outlet that an 8mm barb connector can be screwed into.

RiftAmps

Next job is to remove the thermostat from the front housing and install a new blanking plate, It has the correct size hole to screw in the fan switch.

RiftAmps

Here's the current rear outlet. The green plug is the ECU coolant temperature sender.

Here, the new thermostat housing is installed, along with the thermostat and new gasket. The new housing also includes a place for the temp sender.

And finally, the new thermostat housing cover installed.

strtdv

Looking good. I've just bought an old Eunos Roadster 1.6, I won't be racing it but enjoyed re-reading through the whole thread.

Mine came with an induction kit which I removed and put the stock airbox back on. Changing back to the stock airbox I noticed the car is less eager past 5000rpm and you don't get that addictive sound, but also noticed that the induction kit was causing a flat spot in the power delivery around 3000rpm. 
I reckon a K&N panel filter with the stock airbox is probably the best compromise, and I noticed that BBR GTI recommend a panel filter over an induction kit.

FX_Munkee

strtdv said:

Looking good. I've just bought an old Eunos Roadster 1.6, I won't be racing it but enjoyed re-reading through the whole thread.

Mine came with an induction kit which I removed and put the stock airbox back on. Changing back to the stock airbox I noticed the car is less eager past 5000rpm and you don't get that addictive sound, but also noticed that the induction kit was causing a flat spot in the power delivery around 3000rpm. 
I reckon a K&N panel filter with the stock airbox is probably the best compromise, and I noticed that BBR GTI recommend a panel filter over an induction kit.

A lot of people do wonder why the intake tracts on some of these cars are strangely long, you've discovered why It's to keep a specific amount of air available as it's sucked in. There is a technical term for it, just can't remember it now. I use the stock airbox and a K+N panel filter in mine, as it's the best compromise for road use.

@RiftAmps I swear I said coolant reroute on page one Looks good, nice to see there's proper components available now, it used to require raiding Mazda generic parts and a fair bit of Heath Robinson before.

RiftAmps

FX_Munkee said:

strtdv said:

Looking good. I've just bought an old Eunos Roadster 1.6, I won't be racing it but enjoyed re-reading through the whole thread.

Mine came with an induction kit which I removed and put the stock airbox back on. Changing back to the stock airbox I noticed the car is less eager past 5000rpm and you don't get that addictive sound, but also noticed that the induction kit was causing a flat spot in the power delivery around 3000rpm. 
I reckon a K&N panel filter with the stock airbox is probably the best compromise, and I noticed that BBR GTI recommend a panel filter over an induction kit.

A lot of people do wonder why the intake tracts on some of these cars are strangely long, you've discovered why It's to keep a specific amount of air available as it's sucked in. There is a technical term for it, just can't remember it now. I use the stock airbox and a K+N panel filter in mine, as it's the best compromise for road use.

Helmholtz resonance chamber, helps deal with/control Pressure Wave Harmonics and Intake Tube Pulses

Honda S2000 owners quickly realized that replacing the expertly designed intake pipe with an aftermarket one resulted in a huge loss of power. Sometimes, it's better to not mess with things

FX_Munkee said:

@RiftAmps I swear I said coolant reroute on page one Looks good, nice to see there's proper components available now, it used to require raiding Mazda generic parts and a fair bit of Heath Robinson before.

Yes, yes, I know. We don't actually need it but as I could now remove the HC return pipe from the setup it made sense to do the reroute at the same time. Any excuse to spend money on the car...


LSD news coming soon

FX_Munkee

RiftAmps said:

LSD news coming soon

ooh, I'd immediately say Torsen but judging how you've approached the engine build/s I won't be surprised if you're planning something more exotic.

strtdv

As far as I know if you put the torsen LSD from a 1.8 in, you need the 1.8 propshaft and driveshafts as well

strtdv

Are you using the stock timing or have you advanced it to 14° @RiftAmps ?

RiftAmps

strtdv said:

Are you using the stock timing or have you advanced it to 14° @RiftAmps ?

We’ve got an aftermarket ECU so static timing is set to 10° BTDC and then the map dictates the spark advance for every rpm and throttle position.

No need to go for 14° though on a stock car, it’s a bit of an internet myth when actually 12° is all you need on 95-98RON UK fuel. Those boys in the MX5 cup will all attest to that.