The opinions expressed here are not necessarily the opinions of Allpar.
by Bob Sheaves (2001?)
What is the most efficient car design?
My personal preference depends on the application. For example-in an off road race car- I design (for obvious reasons) a multitube space frame to provide direct load transfer paths to minimize weight and maximize stiffness, to provide a stable base for a long travel suspension to adapt to high frequency, large, variable amplitude waveform inputs.
For similar reasons, in an economy car, I prefer a monocoque [unit-body] to keep weight down and stiffness up (this requires a long travel and generally softer suspension, but more importantly, close attention to NVH isolation) and for a luxury car or truck, a separate body chassis design to allow (in the case of the truck) greater flexibility in designing adaptable vocational bodies and (in the case of a luxury car) easier tuning of NVH and generally lower NVH frequencies to tune out of the design.
The key to efficient design is the intended application - the same ideas vary in cost based on the "how it will be used", so proper design parameters must be defined initially in a new car program to eliminate as must wasted time as possible.
The idea of computers and 3D design programs and FEA, etc. are all nice and can save tremendous time, but, when combined with a lack of project definition, will contribute to waste, instead of eliminating waste.
Lutz pushed responsibility "downstream" as far as it could go, with the QIP (Quality Improvement Process). In simple terms, this meant that management trusted their employees to accept responsibility and do their best to ensure all targets were met.
On the BR (Dodge Ram pickup) there was a "ride and drive" (common term for a management design and performance review at Chelsea Proving Grounds) that decided that the link-coil suspension provided the most acceptible combination of ride, handling, and cost control (meaning it was "cheap") but, when loaded to the design weight (300 lbs passengers, full fuel, and 1/2 payload) looked like hell in that the lower control arm frame mount was visible under the cab where it hung down below the bottom of the rocker and the arm itself ran "uphill" to the axle (the opposite of what looked "good").
To make a long story short- I invoked the QIP system and refused to change it (the key to the ride was that angle being reversed from normal) because it "looked bad". The end result was that my design went into production, because I had the numbers to back up my position. And before anyone thinks that was the "real reason" I left (my intractability causing me to be fired) - this happened 3 years before I quit the corporation. This was one of the reasons Lutz was the popular choice for CEO over anyone else- the man respected responsible individuals. Under Lutz, you were NOT just a "number" or a "body."
One final point - perhaps I was not clear on the timeframe described in my previous post. The time it took to "mess up" (to be polite) Chrysler's profitability was approximately 8 years. The "buildup" started approximately one year after I left the corp under the direction of Eaton. This led to greater and greater "kingdom building" as less and less was directly controlled by Bob Lutz. Eaton usurped the responsibilities of Lutz and gave them to various cronies of his. I lay the entire problem at Eaton's doorstep. I firmly believe that even the Tracinda offer that "led" Chrysler to "merge" with Daimler would have been handled differently under Lutz' control and direction.
The single biggest cost in a vehicle is one thing-labor. Today, in the US, fully 36 percent to 42 percent (depending on the manufacturer) of the labor cost is "non-value added." This means that the labor charged has absolutely nothing to do with the actual design, engineering, and manufacture of that vehicle. This includes "consulting fees" paid to other companies, sales incentives, management costs, sick days, cost subsidized benefits, financial costs (loan points from a bank to pay for everything is one example of this). In other words, the more people you support that do not contribute to the actual vehicles, the more "overhead" you have built into the vehicle.
In the late 1980s and early 90s, when I was with JTE, Chrysler developed into the lowest cost producer because of the efforts of guys like Lutz, Gale, and Stallkamp, all combining to reduce the number of people between them and the guys working in the shops, on the line, and in the design/engineering centers. Then Eaton came in...
An example of this can be seen by asking someone there now, how many management levels there are now between a guy like I used to be (a common vehicle design engineer) and Deiter Zetsche. I can assure you there are more than four, and I haven't worked there for ten years. All of these people cost an industry average of $60,000 per year. Doesn't sound like many people? Consider this, there were 5 groups that used to comprise Chrysler:
.....and in each of these groups the were basically four levels of management, which, as was published years ago in Automotive News, totaled some 750-950 (the numbers varied depending on the source) "non value added" personnel related to the production on the automobiles that were in this cost range. This totals (assuming 850 people @ $60,000/year plus another $30,000 subsidized benefits) $76,500,000 for nothing, IMHO. To put this in perspective, let's look at some other numbers...
Take the total numbers of employees worldwide (let's pick a number of 120,000 as an example only) and only take the US average wage for all of them (yes, I realize that many are less, but management is far more including bonuses) $32,000 plus a subsidized benefit cost of 1/2 that or $16,000 for a total per employee cost of $48,000. The total average labor cost is $5,760,000,000 for the corporation for 1 year....and you haven't built a damn thing yet! An average design cycle is anywhere from 36 months to 60 months (3 years to 5 years) for a total labor cost of (assuming 3 years) is $17,280,000,000. Lotsa zeros, ain't it? Let me say that number another way. The labor charged to the program cost for 3 years of a design cycle is $17.28 BILLION USD.
Now I DO admit that they may have value-otherwise they would not be there, right? WRONG! As was stated, they provide nothing to the corp in the obtaining their main purpose in life-building cars. And they were all gotten rid of under Lutz thru outright firings, layoffs, and buyouts......
Fast forward to today. How many layers are now between the guys that do the job that I do and Germany? Then figure out the costs.
[The] body-chassis vs unit-body (or monocoque, if you prefer) [costs] are quite valid and are a common design review item, but are a small portion (average of 200.00 per vehicle) of the cost, that have much less to do with the overall cost than the labor costs added. Overall sheetmetal "Class A" (the styling you see) costs 4 times that or more.
By the way - government regulations only add 5-7% total to the cost of a car.
Best as always,
All Mopar Car and Truck News