Lifespan of a Single

[Mbirt]

We can all agree that an MX thumper in FSAE isn't going to have the 10-20 years of longevity like a team's legacy CBR600 F2. But it's become painfully clear recently that, for our CRF450X at least, many components not expected to require frequent replacement... do. For the teams that have taken on a single, have you had to reconsider your scheme of rebuilding engines vs. buying another new or used one?

I'm not talking about the things expected to have significantly shorter service life on a single (pistons, valves, etc). I don't mind replacing those once a year. Starters and starter clutches proved themselves to be fragile wile tuning cold and hot starts on EFI. But now our two 2006 motors that have shared dyno and car duty for the '09 and '10 seasons are experiencing more obscure failures.

Mostly in the starting gear train. There's a gear directly before the starter clutch that is pressed together and has a circlip also. The two halves separate, the circlip disintegrates, and the transmission oil is suddenly full of steel (or was that steel left over from the last time the starter sprag clutch blew apart?).

So, knowing that we had two hurt motors, I rebuilt one for the '11 dyno season, replacing all the service pieces and known trouble components. Everything's great the first few days, it's starting on Megasquirt 3, the new big plenum doesn't hurt starting. Let's put the water brake on it.

The starter spins but the engine doesn't. The starter sheared three adjacent teeth from the reduction gear which it drives. Weird. But no problem, there's one on the other motor I can replace it with. I clean the steel particles from entire side of the engine (something I'm used to by now), replace the gear, and throw everything back together. Almost finished with the repair, I notice a 1" long crack emanating rearward horizontally from the orifice which the starter snout sits in. The right crankcase half is now scrap. Excellent.

Kettering's first season implementing the 450X was my freshman year, so I wasn't around in the 4-cylinder days. Serving under the last engine lead, I watched us go to comp carburetted and do the endurance start on 24v twice. The grass just seems so much greener on the other side for this new engine guy that has only dealt with singles.

[dazz]

Hi Mbirt,

We experience similar problems with the Deakin KTM525, I wasn't involved with the team this year (smallest team at FSAE-A 2010 with only 5 members!!) but this is my thoughts on the matter:

The plenum is the problem.
See, when you hit the starter there is a huge volume of air sitting in the plenum at atmospheric pressure. So when the intake valve opens the piston draws air in almost as though you have the throttle wide open - translating to a massive torque required to compress that full cylinder of air. After a few cranks, a vacuum is drawn on the plenum and cranking gets easier. The problem is that the starting system is designed to start the engine with the throttle closed and no plenum - the piston draws some vacuum immediately resulting in much less air entering the cylinder. With only a partially filled cylinder, the starter system has no problems cranking through the compression stroke.

The solution I have thought of is to build a choke butterfly post plenum that can only be actuated for starting - it would have to be built in such a way that satisfies the judges that it IS only a starting choke and CAN NOT be used to throttle the engine while driving (which is illegal). This would allow the engine to start under the same conditions that it was designed to.

You can test my theory:
1- disconnect the intake manifold, ignition & injector
2- crank the engine with an open inlet
3- crank the engine with your hand over the inlet

See which of the two cranks easier and faster...

Hope this helps.

[VFR750R]

Make sure you don't have too much ignition advance for starting. That will kill starter clutches, gears etc in short order even on 4 cylinder models.

5 deg BTDC is realistic

[dazz]

Originally posted by VFR750R:
Make sure you don't have too much ignition advance for starting. That will kill starter clutches, gears etc in short order even on 4 cylinder models.

5 deg BTDC is realistic

Agreed. The key is to compare your system with the stock system and analyse the effect the differences are having. Root cause analysis.

[Mbirt]

Thanks guys!

Dazz,
I'd had your same idea before about a second butterfly close to the motor that better matches the original carb's position. Maybe even a passive spring-loaded valve that only lets your engine see its entire runner length and plenum volume at a flow rates greater than that of full load at the beginning of your powerband. I completely understand what you're saying and the engine does indeed crank much easier/faster when the intake valves aren't seeing ambient MAP.

As a counterpoint, I noticed no decrease in startability going from a 1L plenum to a 3L plenum. Tested back-to-back. The noticeable difference was how it would spike to over 4000 rpm before settling to an idle--most likely due to the larger volume of ambient-pressure air to deplete before creating acceptable vacuum. Oregon state has an interesting video on their youtube channel ("First night on EFI"?) featuring a start on a plenum made of big rubber 5-6" pipe couplings from Home Depot. You can see the plenum contract at least 0.5" during cranking/firing.

The only first-rate team I've seen run a plenum-throttling device in an intake system on a single is ETS. They have a long diffuser from the restrictor that becomes the intake runner to the motor with a 1.5-2L plenum volume branching off it 2/3 the way to the motor. The butterfly sits between the plenum canister and the intake runner. I'm not sure if they only close it off while cranking or if it remains closed during low-mid load/low-mid rpm too, but it's there.

Comments from other successful teams on here running singles seem to hold to the belief that poor starting comes from tuning and not plenum size/geometry. If I had to guess, they're running extremely small throttle settings at idle with tons of advance to keep it running. The tighter seal a the throttle plate should help the motor turn the plenum volume into a good vacuum faster.

VFR,
We've been cranking between -10 and 5 deg BTDC for the last few years. I was disappointed to find that Megasquirt 3 won't let me go lower than -10.

I've also got datalogs showing positive MAP readings (~105 kPa) for a second or two after it fires, spins up, and dies. I don't believe I had any successful starts after a fire that yielded a positive MAP.


My original intent with this thread was to generate a more broad discussion on rebuilding engines vs. buing more new/used ones, however. I'm afraid my team's old method of just rebuilding its few good F4i motors yearly may not suffice with the "tighter-strung" single.

[BrendonD]

Mbirt,

Out of curiousity what range of rpm's are you seeing for decent idle on one of your singles with a plenum installed?

Good thread.

[VFR750R]

Is your megasquirt a 'batch' fire system? Meaning it fires the coil every revolution.

FSAE engines are susceptible to firing the mixture during overlap which will travel back up the intake and pressurize the intake, and it will also burn out all available fuel and oxygen and make it impossible to start until the air/fuel mix have been replenished.

If this is the case, you may be better off advancing your ignition timing a bit to avoid what I call 'front-firing'. You may have to play around with it until you find a happy point that isn't too hard on starting drive-train and starts quickly and consistently.

If you have the ability to alter your injection timing for starting, you may also improve your front fire problems by not having the right air/fuel mix to burn (beyond lean limit) when the coil fires on overlap.

Is it possible to crank a few revolutions without ignition on and then flip your ignition switch? We do that on the dyno with carb engines and they seem to start easier when you do give them ignition. It could be drawing the plenum down before giving it ignition which probably helps with a fixed timing distributor.

[murpia]

Originally posted by VFR750R:
Is it possible to crank a few revolutions without ignition on and then flip your ignition switch? We do that on the dyno with carb engines and they seem to start easier when you do give them ignition.
VFR, do you not think with carbs you also need to build up air velocity to get a sensible air / fuel ratio? Pre-cranking would do that.

Widening the scope of the topic a bit, pre-cranking is sound advice for any modified / motorsport engine. And a good reason to separate starter & ignition switches. It's routine procedure for engines with external starters.

Cranking the engine without ignition will generate some oil flow & oil pressure, scavenge your sump (if you have dry sump) and empty your plenum (if you have one). All are good things.

Of course this might put more strain on your starter motor and battery, so it's important to engineer those right too.

Regards, Ian

[Mbirt]

Originally posted by BrendonD:
Out of curiousity what range of rpm's are you seeing for decent idle on one of your singles with a plenum installed?.

It all depends on what you decide decent is for your team. I've had it as low as 1800rpm by pulling timing out of it. At that speed it's not pulling much vacuum (60+ kPa MAP) and is too easy to kill when shifting out of neutral with the water brake on it. Imagine how much trouble your drivers would have dealing with it. While us engine guys romanticise a low, purring idle, we've decided here at KU that a strong, functional 2800-3000rpm is right for us. Even a hunting idle is better than a weak idle that might let the motor (which is a pain to start in the first place) die.

If your crank/cam triggering scheme is better/yields more resolution, a stronger idle at lower rpm is easier to achieve, I'd guess. We've machined a 24-2 tooth ring that is pressed and tacked to the stock flywheel with the original long/short tooth pattern machined off. It's read by the stock VR sensor. Does your Motec use the stock LTR450 triggering scheme or are you doing something different? How many motors per competition have you guys purchased/built? While you're at it, I wouldn't mind knowing everything Motec told you about implementing their ecu on a single. Haha.


Originally posted by VFR750R:
Is your megasquirt a 'batch' fire system?

We are indeed running wasted spark due to the lack of a cam position sensor. This also means we're squirting the injector every revolution. It's not too late to add a cam sensor, however. MS3 can handle a VR or Hall CPS.

Excellent point about the front-firing burning oxygen out of the cylinder and intake. I hadn't thought of this. This alone could be reason to add a CPS.

Pre-cranking is a good idea for Ian's reasons alone. I can control the coil power seperately from the ecu (ignitor built into COP which receives +5v logic-level signal). So I see 3 possibilities for an "ignition" switch here:
-Switch +12v to ignition coil
-Switch +12v to coil and injector
-Existing ECU power switch
The second choice seems like it would give the most consistent amount of fuel while cranking (imagine having to tune your cranking PW with the first...) while allowing the ecu to get its bearings for crank position better than the third choice.

And anything, Ian, should put less strain on the starter and its gear train than some of the awful-sounding fires and kickbacks this thing does.

[BrendonD]

Originally posted by Mbirt:

It all depends on what you decide decent is for your team. I've had it as low as 1800rpm by pulling timing out of it. At that speed it's not pulling much vacuum (60+ kPa MAP) and is too easy to kill when shifting out of neutral with the water brake on it. Imagine how much trouble your drivers would have dealing with it. While us engine guys romanticise a low, purring idle, we've decided here at KU that a strong, functional 2800-3000rpm is right for us. Even a hunting idle is better than a weak idle that might let the motor (which is a pain to start in the first place) die.

If your crank/cam triggering scheme is better/yields more resolution, a stronger idle at lower rpm is easier to achieve, I'd guess. We've machined a 24-2 tooth ring that is pressed and tacked to the stock flywheel with the original long/short tooth pattern machined off. It's read by the stock VR sensor. Does your Motec use the stock LTR450 triggering scheme or are you doing something different? How many motors per competition have you guys purchased/built? While you're at it, I wouldn't mind knowing everything Motec told you about implementing their ecu on a single. Haha.



Yeah I had it idling fairly healthy at around 2400-2500 rpm, that's with a 2L plenum up top. We were also seeing around 60kPa MAP at that speed. The dyno is only going to give us some room to tune for driveability, I'm sure we'll be screwing with the idle/throttle characteristics once we have it in the car. Part throttle and transients are my priority as of late.

As far as getting MoTeC to run on the single, you basically have to trick it. Tell it that it's running on 4 cylinders, and then just turn off 2, 3, and 4. Once we did that, it started right up. A rudimentary work-around, but this is what we got from one of their reps.

We desperately need a spare engine though... Even if just to get my freshmens' hands dirty.

[Ben K]

Yeah if anyone runs the MoTeC M400 and a single, be sure to do that or the timing gets thrown off and the engine just won't start.

To add to Brendon's comments, the LTR450 engine is fantastic. It already had a dry sump as well as EFI when we got it. Not only that, but it has taken quite a beating and come back for more. ATV engines are pretty glorious.

Ben

[VFR750R]

Could put your 12 volt signal to the coils on a timer? or a relay controlled by the ecu after some code waits x number of revolutions?

I remember I used to have various Jeeps, all with the 4.0L engine and they rotate 6 times before firing every single time unless the key wasn't turned off completely, then they start right back up. I know they had something in the ecm to wait those 6 revolutions.

On the carb stuff, I don't think the boosters do anything at starting. We have to pump the throttle to get shots from the accelerator pumps to add fuel before it will fire, even hot. Seems like priming the engine and then hitting the starter would be better then priming and then cranking without ignition (allowing some of that additional fuel to escape out the exhaust). Trying to start while priming seems to give the worst starting behavior (in this case throttle is opening and closing preventing any manifold vacuum from being created). And as I write this, I just realized the vacuum also evaporates some of the fuel because of the lower pressure which probably makes the mixture more volatile.

[murpia]

Originally posted by VFR750R:
We have to pump the throttle to get shots from the accelerator pumps to add fuel before it will fire, even hot.
If I ever get code any engine management software I'm putting in a 'virtual pump shot' strategy based on the TPS... I miss that cold morning routine of three stabs on the pedal before cranking.

Regards, Ian

[Mbirt]

Northwestern guys,
I assume all the Honda/KTM/Yamaha/Rotax thumper engine guys are too busy rebuilding starters and replacing starter clutches to join this conversation. So that means the LTR450 wins by default. Haha. It's incredible that you guys have been operating on only one motor since '08 though. Showing up to comp with only one CRF motor is unthinkable for us.

I see on your team's website that you used to run megasquirt. What motivated you guys to switch to the M400?

VFR,
I'll see what I can do about a timer. But once I finish my second rebuild of the same engine in 2 months, I'll be trying the switch to the injector and coil first.

Ian,
Megasquirt has an injection priming pulsewidth you can control for when the ecu is switched on. So throw a limit switch on the throttle pedal to kill the ecu at a certain travel, find a way to deactivate it once the engine is running (except for endurance practice when you want to teach your drivers a lesson about fuel conservation, of course), and voila!

[BrendonD]

I wasn't around when the decision to switch was made, but I can say that on the megasquirt, it ran like complete garbage. On MoTeC, it starts every time, idles, and runs pretty decently (working on that). The M400 is a great interface and it's really easy to show someone what's going on without a lot of background explanation. We also have some auxiliary inputs to monitor temps on our waterbrake dyno as well as torque output. (Still needs some signal conditioning on that though.)

We have yet to replace starter components. They're beefy enough I'd guess.

Drawbacks:

It's a new engine, there aren't that many around, so they're expensive to acquire, which is why we don't have more than one at the moment. I remember when we got ours back in 06... it's got some get up and go hidden in there, evidenced by the 0-60 pulls we did in the alley behind the shop when it was still in the ATV. (Fun times... I'm getting nostalgic here)

Not many teams run this engine (probably for the above reason, and they've already built their knowledge base on another model, we didn't have a knowledge base at all when we made a decision). AFAIK, USF and NU are the only teams running one. Again, that's AFAIK, so please feel free to correct me/update my knowledge.

MoTeC is expensive. The Open Source route is probably easier to justify, but for us this works well and I have other projects besides engine management going on, so I'm not going to attempt to fix what isn't broken here.

[BrendonD]

In addition to that, I'm in the process of modifying our dyno to dampen the vibration, cause I'm starting to see early signs of case fatigue from the thump-thump-thumper.

[Drew Price]

Originally posted by BrendonD:
In addition to that, I'm in the process of modifying our dyno to dampen the vibration, cause I'm starting to see early signs of case fatigue from the thump-thump-thumper.


Out of curiosity, how did you come to that conclusion?



I'll chime in on a few things.

-We used MS in in '07 and '08 because we couldn't justify spending more for the benefit we thought we'd see. One member had some experience using MS already, who became the engine lead.

-It didn't run well initially for several reasons, I don't think MS itself was actually any of those reasons.
-No dyno time, almost exclusively road tuned
-Poor understanding of tuning theory (initially anyway)
-Questionable electrical system
-Not enough comfort with the tuning interface
-Poor inlet manifold design. (~500 CC plenum - yeah, 0.5L. It was sloowwwwwwww... we learned a lot about compressible flow after that)
-Ignition system made of junkyard E39 BMW parts


-On reliability
-Never killed the starter / one-way clutch.
-Never killed the clutch
-We had the thing running at around 0 deg. BTDC with no load on it for several minutes at one point, whole exhaust glowing red hot, including 1st chamber of the muffler (titanium downpipe and muffler), and didn't hurt the valves as far as we could tell. Cracked the weld on the muffler, which was where my welding titanium photos came from since we needed to repair it, but I was surprised we didn't cook the engine. That was about the 2nd time we ever had it running on standalone.
-Valves don't need adjusted



-On MoTec
-An NU alum we met is a dealer, gave us a deal on a package and we had some grant money to pay for it. If we were paying for it with earned sponsor dollars we may well have gone a different route, and actually had plans for a while to use MoTec for development and DAQ but to run something less expensive on the car to help the cost report.
-MS has a similar glitch with single cylinder engines, nothing other than Power Commander is really set up to deal with a single.


-On triggering
-The LTR comes with (I think I am remembering this right) a 36 tooth trigger (if you want the actual # the engine lead from my era posted a lot about it here in some of Denny Trimble's threads from Wash.) The trigger tooth is a double-width tooth rather than a missing tooth, which MS did not support, so initially we ground off the double wide tooth to make a 36-1, used stock crank sensor to run MS wasted.

We ran MoTec the same way, and were going to add a cam sync. machined into the cam cover (which is Mag on the LTR, along with both side covers).



-On cost
-LTR's have dropped a bunch in price, especially recently since EVERYONE is selling their toys, there are good deals out there on whole quads.
-I attribute the lack of spare engines for sale to their reliability in stock configuration. Schmooze the Suzi forums a little and you will find lots of people laying 50-55whp on their un-restricted quads with very little about failures.



Drew

[Drew Price]

Post coming, not sure why it was flagged for review.


Drew

[Mbirt]

A quick ebay search yielded several used LT-R450 motors priced in the same $1500-2000 range as the other thumpers we use (except for the Rotax, you'll pay a bit more for one of those). Your lone motor has withstood the implementation of two different brands of ECU on the same outfit of starter and starter gear train parts. I think that's amazing and I'm sure most of the other engine guys on here would agree. Maybe Suzuki, as an early adopter of EFI on a thumper, just equipped their motor with tougher parts?

I'm seeing a bit of a Mac vs. PC analogy to MoTeC vs. Megasquirt. The MoTeC, as the Mac, knows what hardware and inputs it'll see and can provide consistent operation the first time and every time. It's going to take quite a bit more knowledge, blood, sweat, and tears to get the same results out of a Megasquirt setup like this. The savings in stress and time could easily be worth the $3000 price difference. I just downloaded the M400-800 user's manual to see what I can apply to our MS3 setup.

Alright, it's time to get back to the shop and pair some crankcase halves.

[Ben K]

Originally posted by Mbirt:
A quick ebay search yielded several used LT-R450 motors priced in the same $1500-2000 range as the other thumpers we use (except for the Rotax, you'll pay a bit more for one of those). Your lone motor has withstood the implementation of two different brands of ECU on the same outfit of starter and starter gear train parts. I think that's amazing and I'm sure most of the other engine guys on here would agree. Maybe Suzuki, as an early adopter of EFI on a thumper, just equipped their motor with tougher parts?

I'm seeing a bit of a Mac vs. PC analogy to MoTeC vs. Megasquirt. The MoTeC, as the Mac, knows what hardware and inputs it'll see and can provide consistent operation the first time and every time. It's going to take quite a bit more knowledge, blood, sweat, and tears to get the same results out of a Megasquirt setup like this. The savings in stress and time could easily be worth the $3000 price difference. I just downloaded the M400-800 user's manual to see what I can apply to our MS3 setup.


Considering how many times that starter motor was abused when we switched to MoTeC and couldn't figure out that we needed to trick it, you are damn right that the parts have held up well.

Megasquirt doesn't have nearly the options that MoTeC has in terms of data acquisition and such. MoTeC is also run by many of the higher level teams, including GFR, TU Graz, and TU Munich. So it also has one hell of a user base. If you can afford it, I'd get it.

I've heard way too many horror and pain stories about Megasquirt. It'll work, but you might be unhappy with the end results.

Ben

[BrendonD]

TL;DR about MoTeC vs. Megasquirt.

You will screw less up on the MoTeC by virtue of having less things to screw up.

[Mbirt]

Sorry about leaving this discussion for a few days. When not wrenching on the CRF or at my co-op job, I've been on Thumper Talk trying to figure this situation out better.

Thanks for all the info, Drew. My team chose Megasquirt post-comp in '07 when Mahle took their MoTeC back and left us with their 3-cylinder paperweight. The CRF450X was purchased as a desperate plan-B engine in case Mahle didn't deliver their engine before comp in '07. Maybe I'd be running an LT-R on MoTeC too if the team hadn't chosen its future powertrain and controls out of frustration with Mahle.

Aside from starting, Megasquirt's actually been good to us. I've made 44hp at the water brake on an intake runner that was far too long (16.5"--2nd wave at 5000rpm, I do believe) and a small 1L plenum. I took a crack at tuning transients by manually setting the load knob on the brake and rolling into the throttle at various rates. Much to my surprise, it worked and the car was very drivable when the setup was transfered to it. This one particular day last summer is the only stretch of 6 hours in the last two years that we've run completely trouble-free on EFI. I felt every bit of 44hp and it was good. The car ran out of E85 that evening and after we put new fuel in it it just wouldn't start. I can't completely blame the fuel (although it didn't smell as sharply of ethanol as the old E85). It was likely a crank position sensor wiring or coil grounding problem. Nevertheless, we missed the LTU Grand-Prix the next day and ran poorly, but on 100% EFI, at the UofT Shootout.

Back to the present. After pairing the dyno engine's left crankcase half with the race engine's right crankcase half, I got back on the dyno stand and did a few pleasant, trouble-free starts on the carburetor (with MS controlling spark) to get it up to temp and make sure everything was alright. Check. Then came starting on the prototype intake with 3l plenum. After 3 tries with some front-firing on the first two, it started and kept running until I fat-fingered in a 255% VE table entry. The next attempt to start it yielded some more front-firing and then starter geartrain squealing. Great. Please don't let it be the starter clutch.

One EFI start after piecing this motor back together, the "pressed gear," which drives the starter clutch, failed. My research online shows it to be Honda's failsafe in the starter geartrain, albeit a $70-plus one that puts steel particles in the transmission oil. Some posters on various forums have pressed it back together, while other sources say it must be replaced. I've had success re-pressing it once, but it didn't work this time. The starter freewheeled upon installation of the re-pressed gear.

That's enough ventilation on my part. A cam position sensor is needed to prevent front-firing during overlap. I'm also brainstorming ways to prevent the engine from seeing the intake/plenum volume during startup. One way is to go back to a 1l plenum. I'm up to 52hp in GT-Power on a 1l plenum, but it's all near the rev limiter and there's a huge loss in midrange power.

Just in case this thread catches the eye of another team running something other than a 4-cylinder: Do you feel you've found the fatigue life of the major components of your engine? Has it made you reevaluate your engine purchasing scheme? Do you believe buying another new or used motor has saved you money and/or time over doing that second rebuild of an old motor? It would be excellent to have RMIT chime in here.

[ibanezplayer]

Originally posted by Mbirt:
...it started and kept running until I fat-fingered in a 255% VE table entry..

I have fat-fingered many ridiculous things over the years usually killing the engines instantly. I've become a fan of ECUs that have adjustable limits on table entries...

[Kirk Feldkamp]

Another thing of note I've seen on FSAE engines in general is flexible plenums. The ol' alma mater saw it last year on their 600. The prototype plenum they built had a bunch of thin, flat, aluminum sheets welded together. You could easily see the plenum contract at idle, which caused a lot of issues for both starting and staying running after quick off-throttle events. A subsequent redesign for stiffness cured most of the problems they were having up to that point.

Another issue a I've seen a lot of teams struggle with on the singles is injector placement and aiming. On the 600's, it's easy enough to recreate the stock injector mounting geometry, but for (most of) the singles, that simply doesn't exist. Sufficed to say, shooting a stream of fuel directly into a port wall is a poor choice for fueling consistency, but I've seen it more than once. In general, it's best to shoot the fuel directly down the hole(s) and at the back of the valves if possible. Sounds obvious, but I think it's often overlooked.

All I'm trying to say is the EFI setup is an easy target to blame when things don't work, but the design of the rest of the system is just as suspect in many cases.

One last thing... if a huge 255% fat finger occurred and then it wouldn't start again, I'd think that perhaps you wet the spark plug to the point where it wouldn't fire anymore. This sort of thing happens all the time with kart motors after you prime a fuel pump. The cure is usually to remove and dry off the plug, put it back in, and then crank with the throttle wide open until the thing lights off again. There's not enough airspeed to meter any fuel when the carb is wide open, so you're only pumping air in, eventually leaning it out to the point where it will light. For every EFI fuel map I've ever done, I've included a "zero fuel" area at 50%-100% load at the lowest RPM bin (or lowest two depending on how things are set up in relation to cranking RPM). That way, if a driver spins on course (generally resulting in a residual rich condition), he can just put his foot to the floor and crank. The same thing happens as with the kart engine: excess fuel gets exhaled out the exhaust... eventually transitioning to the point where it's lean enough to light off again. This usually works for both hot start and cold-flooded conditions.

-Kirk

[Big Bird]

Hi Mbirt,

I'm by no means an engine expert, but will try to offer some feedback from our side of things.

Agreed that a 20 year life would be unexpected for a 450 single under FSAE conditions. But I think 4-5 years is plenty achievable if you are sensible. And as an aside - sure, the extra mass of a CBR will buy a lot of extra robustness for you, but how many other components of the car get the luxury of the "lets play it safe and make it twice as heavy as it needs to be" philosophy? You rid yourself of a lot of redundant mass by using an engine designed for 55hp - but lose some robustness at the same time. Such is life.

Starter issues - yeah, been there. We were shearing woodruff keys on the crank, an early WR starter problem solved with a Yamaha update part (a spring loaded idle gear between starter motor and crank gear, from memory). It would backfire on starting and just shear the rotor off the crank - occasionally with the rotor "self-machining" a bit more space for itself within the cases. It got a bit expensive before the fix. I don't know what advance we were running, but VFR's advice seems wise.

We've used 3 WR450 engines continuously since 2003, rotated as required. They last saw service for the 2009 car, so 7 cars in total. We've had engine issues at times, but I question sometimes whether the failures were fatigue related, or the result of too many untrained hands fondling the internals.

For 2003 and 2004 the engines were untouched internally, and we never had a problem (aside from starters, above). In 2005, it was decided that we "needed" internal mods (hi-comp pistons, hi-lift cams, etc) and that is when we started to see some major failures. First one - a high comp piston shoved right through the wall of the cylinder when it's skirt dropped off.

We've had some good engine guys through the team since, but they have told of some rather awful handiwork they have found inside those motors. Spacers missing, bolts stripped, cases attempted to be split with screwdrivers and chisels only to find that not all the case bolts had been removed...

These motors can run reliably for the duration of a Paris Dakar Rally or an Oz Safari, and still come back for more. The trick is to treat them well.
- Train a proper engine builder on the team, and follow proper process (assembly done exactly to factory spec and process, every bolt tightened to torque and duly marked, everything inspected and measured where possible, correct tool for job every time, bench kept spotless, NO-ONE aside from the builder touches any component of the open motor etc). Speak to professional engine builders if necessary - listen to what they say.
- Don't try to overcompensate for the smaller motor. You've chosen a small engine, it has advantages (size, weight) and some perceived disadvantages (less power). Play to the advantages and don't fixate on the disadvantages. At best, engine mods might get 2-3hp increase somewhere in the range. At what risk? If you really "need" more than that - then you shouldn't have chosen such a small motor in the first place.
- Obvious, but use good quality oil and change it regularly.
- And if you must change stuff, do it one thing at a time. And in small increments.

I hope the above helps and isn't too sermony. I think you can get a good run out of such things if you do it properly. Good luck with it all.

Cheers

[Big Bird]

Just in case this thread catches the eye of another team running something other than a 4-cylinder: Do you feel you've found the fatigue life of the major components of your engine? Has it made you reevaluate your engine purchasing scheme? Do you believe buying another new or used motor has saved you money and/or time over doing that second rebuild of an old motor? It would be excellent to have RMIT chime in here.

Sorry, should specifically address the questions!

Q. Do you feel you've found the fatigue life of the major components of your engine?
A. We've found the engine life limit after 7 years and about 12-13 comps (3 engines) - that may be fatigue related, but I believe a lot of it is due to imperfect work practice.

Q. Has it made you re-evaluate your engine purchasing scheme?
A. Yes, we have finally bought new engines this year.

Q. Do you believe buying another new or used motor has saved you money and/or time over doing that second rebuild of an old motor?
A. We were onto 4th or 5th rebuilds of old motors. Many times we have argued that the team needed new motors in the past, but the money was redirected to other areas of the car. The engine reliability issue was noted when it started impacting on results.

I hope the above are specific enough, I can't really offer too much more.

Cheers,

[Drew Price]

A. We were onto 4th or 5th rebuilds of old motors. Many times we have argued that the team needed new motors in the past, but the money was redirected to other areas of the car. The engine reliability issue was noted when it started impacting on results.





Keep in mind than when rebuilding singles or twins you can often get NEW cylinders to replace along with new pistons, rather than having to bore / sleeve an inline block.

That along with touch-up work on the crank and new bearings might actually land you with a more 'complete' rebuild than what many people I talk to (consumers and racer types) perceive to be a rebuild, which is often a 'Rings & Bearings' rebuild, which is what I would call a 'Refresh,' not a rebuild.

Sometimes when my shop owner is nearby I call this a 'Peruvian Rebuild'.....

An actual rebuild will usually have you replacing (or at least measuring) the timing components as well.


Drew

[Demon Of Speed]

Originally posted by Big Bird:

These motors can run reliably for the duration of a Paris Dakar Rally or an Oz Safari, and still come back for more. The trick is to treat them well.
- Train a proper engine builder on the team, and follow proper process (assembly done exactly to factory spec and process, every bolt tightened to torque and duly marked, everything inspected and measured where possible, correct tool for job every time, bench kept spotless, NO-ONE aside from the builder touches any component of the open motor etc). Speak to professional engine builders if necessary - listen to what they say.
- Don't try to overcompensate for the smaller motor. You've chosen a small engine, it has advantages (size, weight) and some perceived disadvantages (less power). Play to the advantages and don't fixate on the disadvantages. At best, engine mods might get 2-3hp increase somewhere in the range. At what risk? If you really "need" more than that - then you shouldn't have chosen such a small motor in the first place.
- Obvious, but use good quality oil and change it regularly.
- And if you must change stuff, do it one thing at a time. And in small increments.

I hope the above helps and isn't too sermony. I think you can get a good run out of such things if you do it properly. Good luck with it all.

Cheers

MBirt, read this!

[Mbirt]

Sorry DOS, I created a thread and asked a bunch of questions without a single drop of intent for reading anyone's responses. Especially the responses from a respected and knowledgeable member of a team which I requested a response from. Thanks for bringing it to my attention that I should read the response to a question which I asked.

I do need to pick an engine builder. A grad student with no class load might be perfect for this job. The powertrain design and calibration guy is taking 20 credits this term...

[Mumpitz]

Engine builder: better have grease perma-stained into their hands and an arm that comes with a calibration cert.
Joking aside free time and class standing is the last qualifications to look for. It doesn't matter how busy or how much of scholastic over/under-achiever the person. They need to have the patience dedication and appreciation for the task at hand. The person with the right qualifications will find the time.
If you are dealing with engines that have seen wear and tear it is extremely important to have someone who knows what they are looking at and can spot abnormalities from a mile away, not just follow steps in a service manual.

[Mbirt]

Hey Mumpitz! Do you go to OU? How's the transition to MS3 from MoTeC going?

About one's qualifications to be an engine builder, I was just joking with DOS. He's trying to fly incognito on here, yet knows about my situation well enough to choose a highly situationally appropriate quote. I find that creepy. But, fortunately, I've been creeping on him too. He's the grad student taking no classes next term.

It's been nagging me for about a year that I need help that deals with only the engine. The engine has been trouble so many times in the last 3 years that almost every team member has helped pull the head, rebuilt a starer, or timed it when everything's gone back together. So, while everyone knows a little bit about the engine, no one is competent or experienced enough to fully rebuild one. The last engine guy even left the stock valve springs in when he replaced the stock titanium intake valves with stainless ones. Good thing I made a trip to a local thumper builder who was able to point this out to me.

It's time to recruit and retain some engine-building talent. I'm going to make some sweet posters for the start of this new academic term.

[Mumpitz]

DOS, so you're taking a term off to chase Ernst's? That's what I'm guessing.

Yep, John from OU. MS3 is goin, slow progress though. Still waiting on some parts. Ordering things through our school is the most awful drawn out process known to man. The guy we bought engines from said two schools bought and paid for engines in the time it took our school to cut a check. Come spring we'll have to have a test and tune day out here. You bring the brews, I'll build you a muffler http://fsae.com/groupee_common/emoticons/icon_wink.gif

[Demon Of Speed]

For the record. I am taking a full grad load (8 credits). One class is "advanced IC engines" and the other is an independent study which involves me getting hands on learning with engine turning and calibrations (and everything else that goes along with that). As I am a DV guy, this should be new/fun stuff to learn but I don't know a lot about what I will be doing.

Also MBirt, I know you were going to read this thread, just thought I would highlight some information I thought was good.

[Mbirt]

Originally posted by Mumpitz:
Come spring we'll have to have a test and tune day out here. You bring the brews, I'll build you a muffler http://fsae.com/groupee_common/emoticons/icon_wink.gif

Haha. DOS is the infamous muffler killer. Our mufflers are going to be under intense scrutiny during pre-comp testing. I refuse to let the muffler be the reason for an endurance DNF.

A test and tune day at OU would be incredible. I attended an Alfa club auto-x during the summer of '09 and the huge lot there was great. But I was disheartened to see it being torn apart for construction around the time of the Concours this summer. How's the construction going? We're going to Cali this year, but my drop-dead date for EFI in the '09 car is April 1. So, even if our '11 car isn't driving, I'll have an engine team and the old car there to do calibration work with.

I'm going to PM you some more muffler details.

[Mbirt]

Closer to the original topic, it seems that the lifespan of a single for UC Berkeley was 2 years:

fsae . berkeley . edu / cars . html

They ran the CRF450X with a MoTeC M4 in 2008 and 2009, before moving back to their GSXR600 in 2010. Their news update for September 2009 states, "This year we are planning to use the same monocoque as last year, but rebuild the rear frame around a four-cylinder GSX-R600 engine. This engine has a stock fuel injection system, is easier to tune, and can produce almost twice the horsepower of the single cylinder Honda CRF-450X we used last year (~70 vs. ~40)."

They did what they felt was right for their team and showed a significant points improvement in 2010 to back their decision. Does anyone know more about what went into UC Berkeley's big decision post-comp in '09?