Quote:
Originally Posted by bruce.augenstein@comcast.
Because of the second reason you do this, which is that you can build a better engine through hand assemby.
When you're machining various engine parts prior to final assembly, there will inevitably be variances in the final product. Take a cylinder block, for example. Boring each cylinder heats the block in that area, and depending on cylinder placement and the volume of adjacent material, the amount of heating will vary. Then, boring another cylinder starts with a hotter block, and so on.
In addition, once you've bored your hundredth block with the same cutter(s), they have worn a bit.
And so on.
Let's say you call for five thousandths of an inch piston clearance, plus or minus a thousandth, in the build process.
Hand assembly can result in an engine that has five thousandths of an inch clearance on every cylinder, plus or minus nothing, because pistons have tolerances as well, so "oversize" pistons can be mated to oversize cylinders, and so on.
Such an engine will tend to make a bit more power, run more smoothly, and perhaps even get slightly better mileage.
50 years ago, an engine like this would've been termed "a freak" in the parlance of the day, resulting from the chance assembly of exactly the right piston for each cylinder, etc., creating more power, running more smoothy, and so on.
Nowadays, overall tolerances are tighter, so there is less engine-to-engine variation. But there still are tolerances in the process, so if you take ten automatic M3s selected at random off the assembly line, you will still get ten different ETs and speeds over a quarter mile, even if each is using launch control. I'd venture to guess two to three tenths variation in ET, and perhaps as much as a two mph difference in trap speed. (Note that only some of this variation comes from differences in engine power. The rest can be attributed to overall build tolerances in the drivetrain with it's associated bearing clearances, etc.)
If those M3s had hand-assembled engines, there would tend to be less variation, and the results would be clustered closer to the quick and fast end of the spectrum.
So to recap, hand assembly will tend to result in an overall "better" engine.
|
Although it probably cannot be settled here among us I would take the completely opposite stance. No surprise there I guess... I agree with you in the first part of your post but you go awry later on in the parts I have quoted.
Things built by machine are in general more accurate, more consistent and these issues contribute greatly to overall quality.
One also should not confuse the machining and grinding accuracies and tolerances of individual components with the assembly of parts manufactured using these processes.
In modern automotive engine machining and grinding tolerances can be measures in microns or even smaller. Higher quality engines will have improved tolerances. Modern statistical process control and 6 sigma manufacturing techniques along with 100% in process part inspection compensate very well for items like tool wear, heat, etc. These are nearly non-issues. Of course higher accuracy and smaller tolerances still cost more money to manufacture and higher end engine components will have better controls in place for this compared to those going into cheap engines. But then again the cheap engines of today have components more consistent and accurate than higher end engines of just a decade or two ago.
Now again we must not confuse component accuracy with assembly quality. They are not all that related. Volume of production is the almost the sole factor that will determine how much of an engine is hand built (hand assembled to be clear since ALL parts are CNC machined). For low volumes the investment in very expensive and accurate robotics for assembly simply does not provide a return on investment. Just hire the skilled labor and even in small numbers they can meet the production requirements.
Just about everything a human can do with his/her hands a robot and computer can do better. Now that being said it also requires a very ingenious machine designer and engineers to design the assembly machines In this endeavor as in any the consistency, speed and flexibility of these machines can vary widely. You pretty much will get what you pay for. The thing is that this field of manufacturing engineering is extremely mature. The things that assembly by such systems prevent are more of the gross human errors like banging and damaging precise parts, skipping assembly steps, skipping a small component, mistorqing a fastener, slightly damaging a seal, etc.. Modern machine assembly will generally exhibit many fewer of these types of problems than human assembly thus improving overall quality.
I believe our misbeliefs about this human vs. machine thing come from some expired and romantic view of human superiority over machines. 40 or so years ago this may have still had some validity but today it just doesn't. Surely there still are (not much in the automotive world though) bad/crude assembly machinery and robots that can be outperformed by a carefully trained, managed and monitored group of humans. This just is not the norm at all.
- Automated manual transmissions are simply better, faster, more accurate and offer huge performance gains over manual transmissions (thanks to "robots" and software).
- Modern engine ECUs and direct injection are way better than carburetors (thanks to computers and electronics)
- Airbags and active shoulder harnesses are way better than pendulum operated lap belts (thanks computers, electronics and MEMS sensors).
And similarly machine assembly is generally superior to hand assembly. Can you even imagine what a totally hand assembled or hand built car would look like? I would not want to drive one.
I completely agree that MB's claims about this engine are a huge stretch of any reasonable definition of "hand built" and in doing so they are simply continuing to promote a curmudgeonly/antiquated view of the superiority of robots and computers for the vast majority of high volume manufacturing.