The Joy Of Owning Two Old German Cars Is Having Both Of Them Crap Out On You In The Same Week

My BMW had let me down, throwing off its crank pulley in hilarious fashion at 50 mph. “No matter,” I said. “I’ll just drive my Mercedes until all this blows over.” So, champ, how’s that going for you?

Not well. I punted around in my 1998 Mercedes E240 for a week or so while I decided on a plan of action for the BMW. I’d sourced the cheapest crank pulley I could find—$120 AUD from Maxspeedingrods—and waited upon shipping. And yet, before I could deal with it came a warning, unwelcome. “COOLANT LEVEL” screeched the dash of the Mercedes. “Okay, maybe just a top-up needed,” I mused.

No such luck. The next day, the same, with telltale spotting on the ground to boot. It was losing coolant fast enough to be a problem—my problem. I had two cars, and neither one would get me down the road. Two cars down in as many days! The shame of it all.

When I became the lead host of SAUSAGES Und SEDANS (2023), I never thought I'd own two dead German luxury four-doors at the same time pic.twitter.com/RefticVZ5F

— Lewin S. Day (@rainbowdefault) February 5, 2024

Worse, I felt powerless to deal with the problem. With the BMW, I was stuck waiting for parts to ship. With the Mercedes, I couldn’t even start to investigate. Some fool decided to landscape this place with chunky, ugly gravel. It makes even the simplest job an utter pain in the ass. If you’re gentle, you can maybe get down on to your back, and wiggle your way around on the pointy rocks, for about three minutes at a time. Then you explode in rage at the irritation of thousands of tiny points jabbing in your back, neck, and head. Even better when they’ve been heated to a nice roasting temperature by the giant cancer lamp Australia calls a sun.

As best I could tell, the Mercedes was either leaking from the radiator, or close to it. But I couldn’t get under the car well enough to see, and I couldn’t get it on ramps because of the rocks. I couldn’t just yank the radiator, either. By virtue of being an automatic model, it has ATF lines running to the radiator that need to be disconnected from under the car. If I wanted to fix this car, I’d have to dig.

In a week with temperatures dancing around 100 F, I set about clearing a usable driveway. In the past, my tires had dug through the gravel to brick below, and I’d hoped that perhaps some remnants of a sane person’s driveway lurked beneath. A few heavy shovelfuls revealed my hunch to be right, and I started digging in earnest. It was hot, sweaty work, particularly for someone that spends 40 to 60 hours a week sitting on a chair for a living (sad, I know). But I persevered, and carved out a dusty, filthy little patch on which I could repair my car.

I set out my ramps, and tried to get the Mercedes in the air. By virtue of the slope of the drive, or maybe the heat affecting my judgement, my first attempt was a poor one. I overshot the end stop on the ramps, which ended up under the sills instead. A lovely benefit of owning plastic ramps is that this embarrassing faux pas of mine left no real damage. Round two, I got the car up just fine. I shored things up by building a gravel berm in front of the rear tire to ensure the car wouldn’t go anywhere.

I then proceeded to fight the Mercedes for the radiator’s hand in marriage, or whatever. At first, it seemed like it would be easy. The top radiator support actually unbolts from the body of the car, no problem at all. “It’s like working on a tube-framed S13!” I chuckled, thinking this would be easy. But alas, no.

I found myself laying on the brick driveway getting as filthy as a rat that doesn’t just eat garbage, but loves garbage. See, when you lay gravel, it turns out, it’s typical to lay down a layer of sand or other softer material underneath to give it something to sit in and grip on to. That thick layer of dust was getting all over me and in me, and it was barely any better than working in a mud bath as coolant and ATF rained down from above.

Oh, yeah, fun note. The coolant level warning comes up when you lose a liter or two (a quart or two, okay) of fluid from the overflow tank. The cooling system actually still had six or seven liters (six or seven quarts, keep up) inside, at least until it started spewing it all over dusty little me. My new red brick driveway was now covered in shades of green and brown.

Beyond the hoses, I had to pull off front trim pieces, the electric fans, the airbox, a supplementary air duct, and I had to unbolt the air conditioning condenser, too. It seems Mercedes thought that a W210 having a radiator problem was pretty unlikely. Pretty bold stuff from whoever chose to spec out plastic end tanks.

After pulling apart what felt like half the front end, I finally wrested the radiator free. Only for it to appear largely intact. There were some marks on the sides where it appeared coolant had leaked, but I wanted a smoking gun. I needed to know that this was the part at fault, because it was proving incredibly difficult to find a replacement for less than $500.  The marks on the driveway suggested it was leaking from the passenger-side end tank, but I wanted to be completely certain.

seems like end tanks pic.twitter.com/EPlJY4toml

— Lewin S. Day (@rainbowdefault) February 12, 2024

I ended up plugging all the outlets on the radiator and filling it up with water, praying that I could find a leak. I wasn’t heaps confident, because my garden hose in no way recreates the heat and pressure that the cooling system sees in use. It was possible that the end tank was only leaking when the engine was hot, which would make this a boondoggle. In the end, though, I found a small deformation in the end tank, a sort of dent, that seemed to be weeping slowly. I decided that was proof enough and elected to buy a new radiator.

I would have tried to source one from a local wrecker but E240s don’t come up all that often. Instead, I redoubled my efforts and found a decent candidate for a few hundred bucks. I’m waiting for that to ship now.

In amongst all this, though, I found time to enjoy a nice hot dog and moan about my car woes. I shot another episode of Sausages Und Sedans, the Internet’s only sausage car show. The Mercedes was the star, along with the Debrezener sausage from Hungary.

You've waited so patiently. The team and I are so excited to bring it to you.

Finally. It's Episode 2 of SAUSAGES und SEDANS!

Tell your cousin
–we eatin' good. pic.twitter.com/zYzz7H8CPQ

— Lewin S. Day (@rainbowdefault) February 8, 2024

Bimmer Time

In the meantime, I’d had parts for the BMW finally show up. A new crank pulley, what joy! It attaches with four bolts, so it surely had to be an easy swap, right?

Now all I had to do was change it over. I’d elected to go for a straight swap in the interest of saving money. My existing accessory belt and tensioner seemed fine, with no real wear or noise in the bearings to speak of. My cost saving ethos wasn’t all good, though. My cheap replacement pulley—just $120—had not come with replacement bolts.

Why was this a problem? Well, certain bolts are designed as “torque to yield” fasteners. To understand that, we need to understand materials science. The very simplified explanation is that you can take something like a steel bolt, and stretch it. This is what happens when you tighten a fastener—it stretches a little bit longer, often imperceptibly so, as you’re tightening it down. If you do this a small amount, the deformation to the bolt remains in the “elastic” region, where the bolt will return to its original size when you untighten it. In contrast, torque-to-yield fasteners are tightened past the yield point of the material, and they are stretched into “plastic deformation.” When stretched to this level, the fastener does not return to its original size when loosened, and should be replaced after a single use. Torque-to-yield fasteners are chosen as the technique generates a higher preload on the bolt and increases its fatigue resistance, useful in high-stress applications like head bolts, for example.

It’s not absolutely clear to me whether the crankshaft pulley bolts are torque-to-yield fasteners. The M10 bolts have a torque spec of 40 Newton-meters, plus a further 120 degrees of rotation. This kind of spec is at times used for torque-to-yield fasteners. However, it’s also sometimes used as a more reliable way of ensuring a fastener is torqued properly, by ensuring a baseline plus some extra rotation. BMW seems to specify that they be replaced after a single use, but they do say that about a lot of things.

Okay, real talk. I’m supposed to be an engineer. We’re supposed to be able to answer questions like this using maths and paper, so that’s what I did. I wanted to find out whether it was likely these bolts were being torqued to the yield point of the material, and what the possible final torque value actually was. So I did a lot of maths using these formulas. This gets heavy; feel free to gloss over if you just want the answers.

I made some assumptions that engineers will like to look over (k value of 0.2, lubrication factor of 20%, material of the bolts). I also based stress calcs off the minor diameter of an M10 bolt (8.97 mm), but did the torque calcs based on nominal diameter of 10 mm. I also may have made some major mistake somewhere; some of these numbers felt a little off. Take everything here with a grain of salt.

Anyway, based on these numbers, I estimate that the initial 40 Nm torque stretched the bolt by 2.27 mm, or roughly 0.09 inches. That sounds excessive to me. But then, based on the thread pitch of an M10 bolt, we can calculate that 120 degrees of rotation would stretch the bolt by a further 0.5 mm (~0.02 inches). So, for a total stretch of 2.77 mm (0.11 inches), I calculated the bolt was at a stress of around 554 MPa, with a final torque application of around 48.8 Nm.

There’s quite some stretch involved, to be sure. At the same time, that total stress of 554 MPa? Well, let’s look at that. If the crankshaft pulley bolts are say, Grade 8.8, medium carbon steel? They have a yield strength of 640 MPa. If a couple of my numbers shift, and the stress is a bit higher than I calculated, then yeah, maybe these fasteners were torqued into the yield region. But if the bolts are, say, Grade 10.9, which could be likely for a crankshaft-related fastener? The yield strength of that material is 940 MPa, and it’s unlikely the fasteners were plastically (permanently) deformed by the first tightening cycle.

Ultimately, all this math is pointless because I don’t actually know what metal the fasteners are made of. I did all the maths after the fact to see if I could justify reusing the fasteners. Really I should have just spent the extra money, but who has the time to wait for shipping?

I elected to try my luck reusing the bolts. But first, I had to remove the damaged pulley. This looked easy enough, but the confines of the engine bay conspired against me. I could just about wedge my rattle gun onto the bolts, but it was jammed up against the radiator fans, leaving no room to actually back them out.

drill is stuck in the beemer, send help pic.twitter.com/Wu5dF6UWhG

— Lewin S. Day (@rainbowdefault) February 12, 2024

Instead, I had to crack them loose using the rattle gun, snug them back up gently, and then use hand tools to remove them. With the fasteners removed, the old pulley came straight off, and I even found a bonus chunk of rubber that had fallen off when the part failed. Bonus!

The new pulley slid right on, no problem. Tightening it up would prove more difficult, however. The initial part of the torque spec was to tighten all four fasteners to 40 Nm. Here’s the problem—the crankshaft moves. I’ve got an auto, and so there’s no way to jam the car in gear to lock out the crankshaft. You’re supposed to use a special tool to lock the crankshaft in place for this job, but, you guessed it, I didn’t have one. I found a guide online that explained how to use a borescope and an Allen key to lock out the crank at the back of the engine, but lack of access again seemed to make it impossible.

I re-watched a number of videos online, one of which explained that fitting the accessory belt can be done before fully torquing down the pulley. I tried that. I was able to get to about 33 Nm on each fastener reliably enough, tightening each one as the engine was on a compression stroke, which provided some resistance to rotation. I then did another pass and I reckon I probably got most, if not all four, up to 40 Nm. Ish.

But what of the further 120 degrees of rotation required? Well, my inspiration here was straight out of the YouTube comments section. Simply use the rattle gun and watch the fastener as it rotates. I reckon I got about 90-120 degrees on each one. It probably helped I was using a semi-flat battery in the impact wrench. Fully charged it might have over done it, snapping the fasteners.

OH YES

beemer status: operational! pic.twitter.com/iGHQg4SnUl

— Lewin S. Day (@rainbowdefault) February 12, 2024

Starting the car, the engine burbled into life just fine, with the accessory drive neither self-destructing nor wailing a complaint. Generally, anything spinning at thousands of RPM will make you aware in short order if it’s unhappy. I’ve taken it on a few drives and it all seems to be happy enough.

But what of the bolts? Well, if they truly were fragile torque-to-yield bolts, I could have trouble in short order. I’ll be keeping a close eye on the crank pulley, making sure they stay torqued up. The math has me quietly confident they will. Worst case I can order some new fasteners and fix my work. Even then, I’d probably want a crank-locking tool to do the job properly, and that could be difficult to find or expensive.

Anyway, I drove around and parked my car somewhere to take photos, just to prove to you all that the Bimmer actually runs. And you know, so old mate Peter has something to use for a topshot.

For now, though, my fleet is back to 50 percent readiness. I just need that radiator to come in and I can get the Mercedes back together, too. Let’s see what other sick surprises my German sedans have in store, though. I suspect my efforts to sell up and ship out could yet have further complications.

Image credits: Lewin Day