Keep on Rolling (Post # 111)

The abrasions on the inside of the Borrani wire wheel rims
show as dots where the brake drums come closest to the
wheel rims. 

The front brake drum profile. Note the four distinct facets

that mimic contours inside the wheel rim. The "furrow"

 shows on the fins of the middle facet.

After putting about 5 km on The Alfatross in preparation for the Arizona Concours d'Elegance I noticed some abrasions on the inside of the front Borrani wheel rims that seemed to be the result of the rims lightly touching the finned drums in some places. There was no interference problems between the rear drums and wheels because the drums there are much smaller in diameter. 

I checked the fins on the drums and observed that every fin on one of the four facets on the perimeter of the drums has a shallow “furrow” in it corresponding to the place on the wheel where the abrasions appear. But the furrow was not caused by contact with the wheel.  It appears to have been evenly and professionally machined into the fins before I got the car, so it may even be a “pre-existing condition”.  


But what was causing the problem and how do I fix it? Time to contact the Alfa 1900 Brain Trust members and Cork Adams, the guy who restored the wheels, for advice. 

Cork opined in an e-mail "It appears to be a wheel that has slightly too much positive off-set and might require some minor adjustment to pull the hub inward. I assume a quarter of an inch should not move the rim outward enough to rub on the fender." (Groan).

Giuseppe Maranghi suggested that I test-fit the car's other wheels to see if perhaps the wheels and drums are supposed to be paired specifically  If not, I should get in contact with Matteo Bosisio at Borrani to get his take on what the problem might be--after all, they are literally the ones who "invented the wheel"!  Then Joost Gompels sent me what appears to be a period Borrani shop drawing with dimensions so I could check my wheels against it and added hopefully, "I will be interested in what you find out."

Design drawing with dimensions for a Borrani wire wheel .  Joost Gompels.

Suddenly, not only did I have a new, totally unanticipated problem, but it was starting to look like it could take a lot of time and money to fix. 

It seemed like the advice and observations I was getting boiled down to only three possible causes:

1. The wheel hub is centered too far to the outside of the wheel rim. If so, the wheels were assembled wrong during their restoration.
2. Some kind of spacer or shim is missing somewhere on the hub. But there aren't any parts left over or any reference to shims in the design drawings and no place to add a spacer anyway. 
3. Wheel hub isn’t tight enough on the spindle. But how do you estimate how much torque you are putting on the knock-off "wing nuts" when all you can do is hit them with a hammer?

The splined spindle that mates with the hub of the
Borrani wire wheel.  Note the dark abrasion around the
tapered base of the spindle indicating a loose fit.

Because possible cause no. 3 was the easiest to address, I decided to start there, just looking for things that . . . didn't look right. Abrasions on the splined front wheel spindles matched wear marks on the insides of the wheel hubs where they made contact, suggesting that they weren't mating properly, so I sanded and polished both surfaces until I was satisfied that all burrs and refinishing residues were removed.  

But the question of how do you achieve the optimum wing-nut torque, preferably without beating them with a lead hammer, remained.  Beating with a hammer--how barbaric! I used to think that you could find the answer to any question on the Internet if you spent enough time.  I don't anymore.  Try to find the "correct" torque values for knock-off wing nuts on line.  You find everything from "hand tight and then 3 good whacks with the hammer" to "250 to 350 Nm" (184.39-258.15 ft/lb). In the end, I decided to just come up with something I could use that would get the damn wheels seated good and tight on their spindles without hammering the wing nuts to death in the process.  

Precisely at that moment, my neighbor Wyatt Fenn showed up with a 6-pack of pretty good micro brew and time to kill.  As I told my predicament while crying in one of his beers he mentioned that he used to be in prototype manufacturing and promptly sketched a simple wing-nut wrench that could be made of  no. 810-A7 nylon stock. Uh, sorry, Wyatt I'm fresh out of 810--but I do have some heavy plywood . . . .  No problem, plywood works too! 

Rube Goldberg "Proof of concept" wing nut wrench.  Note the finely-crafted plywood workmanship, the profligate use of sheetrock screws, the hand-routed recess for the wing nut, and the clever hole in one arm for easy hanging storage.  Wyatt Fenn design.

A couple of hours later the 6-pack was gone and we had our "proof of concept" wing nut wrench.  I used it for several days tightening and loosening the wheels and it proved three things:  First, that it was vastly superior to the hammer technique and, along with cleaning the mating surfaces, solved the problem completely; second, that all it cost was Wyatt's 6-pack and third, that it could be better.  

More searching on the Internet led eventually to Herr Michael Kirchgassler at Oldtimertools (mk@oldtimertools.de) and his product FMB 42-2. It accomplished the same things that Wyatt's proof of concept did, but more elegantly.  Its compact size eliminated the problems created by the long arms of the proof-of-concept design having only a few degrees of movement before contacting the car body, and its hard plastic wing-gripping toggles were even less injurious to the wing nuts.

The Oldtimertools wing nut socket: What it lacks in 
affordability it makes up for in beautifully engineered 
German manufacturing.   No more guessing at how much 
torque you put on your wing nuts!  

So in the end it was a relatively easy problem to solve. The front wheels were restored properly and no shims or spacers were needed.  The wheels just weren't tight enough on their spindles.  A more experienced restorer could have figured that out in a heartbeat,  but in the end I sure learned a lot and acquired new respect for the deceptively simple-looking wire wheel--and the indispensability of a prototype manufacturing neighbor with a 6-pack of pretty good microbrew!

Weights and Measures (Post # 110)

How does The Alfatross compare to its siblings and other cars of the period, following restoration?  Did it gain or lose any weight, change its dimensions or balance?  Does it run as well as it did when new--or better?  Well, it's early days yet, but we can answer some of these questions.

To put things in the proper perspective we need to start with an historical review of the different variants of the 1900 series. The 1900 model type started in 1950 with 5 cars, chassis numbers 1900 00001- 1900 00006, all Berlinas (enclosed 4-door sedans). Powered by the model 1306 1884 cc engine. 

The next year Alfa added the Sprint model (1st Series) to their lineup but only 6 examples were made.  These were sportier with a shortened chassis and more powerful engine, the 1308 1975 cc engine.

The 4-wheel drive "Matta" version incorporating
Alfa 1900 mechanicals. Millenove, H.-J. Dohren.

Between its original appearance in 1950 and the end of production in '59, a total of 21,304 1900s were built. Naturally, over such a long production run, there were many modifications and variants, including sub-models 1900C, 1900L, AR51, TI, and AR52.  The AR models were "Jeepish" military utility vehicles with 1900 DNA, but not on the same limb of the family tree as the road cars.

The Alfatross was created on a 1900C chassis in 1955.  In that year Alfa produced four variants, the 1900 Super, 1900 TI Super, 1900 Sprint 2nd Series, and 1900 Super Sprint. The C suffix stood for "corto"--the short, 2 door chassis--and was stamped into the firewall after the 1900 designation.

Chassis *01955* was originally created with a flat roof.
Now it has a double bubble roof.  John de Boer.

How Rare is The Alfatross?

According to Gonzalo Alvarez Garcia's research in his book Alfa Romeo 1900 Sprint, as a sub-set of overall Alfa 1900 production, the Zagato-bodied cars turned out in 1955 is microscopic:

21,304=Total type 1900 
854 =Total Super Sprint 
248= Total Super Sprint in 1955  
14= Total 1955 Sprint and Super Sprint Zagatos

The engine of chassis *01955* is extensively modified.
John de Boer.

Of the original 14 Sprint and SS Zagatos two of them, chassis numbers *01964* and *10095*, seem to have disappeared.  The remaining 12 have been or are being restored. Photographs are available at http://www.alfacompanion.com/alfa1900/stories2/extant_zagatos/index.html.  


Although they were produced in the same year, each of these cars is different.  After 61 years of use, abuse, and "personalization" the differences between them are even greater.  One car, *01955*, started life with a "normal" flat roof, but was given a "double bubble" roof treatment later.

Performance Check

The 1900 owner's manual gives the following performance figures for maximum horsepower at corresponding rpm (no torque values are offered): 

1900 Super 90 HP @ 4000 rpm1900 TI Super 115 HP @ 5000 rpm1900 Super Sprint 115 HP @ 5000 rpm

Following the rebuild of The Alfatross' original engine, during which new cams, pistons, valves, guides, and rods were produced, horsepower and torque figures improved significantly.  

The horsepower and torque figures for The Alfatross as determined during several dyno runs.  

Longacre Accuset computer scales model no. 72594.

Weights and Measures

I wonder how many of The Alfatross' siblings know how much they weigh or how that weight is distributed? According to the owner's manual, an Alfa 1900C SS coupe in running order should weigh 1000 kg. (2200 lb) while the cabriolet should weigh 1150 kg. (2530 lb)  No figures are given for weight distribution.

After weighing The Alfatross "in running order" with about one quarter tank of gas using a Longacre Accuset 4-wheel scale we came up with a total of 1002 kg. (2204 lb).  This is puzzling because we expected it to be considerably less than the Alfa 1900C SS with its production body and interior. We weighed the car several times with the same result.  

There were other surprises as well.  The weight distribution was a nose-heavy 59% front (595 kg./1308 lb) vs. 41% rear (407 kg./895 lb).  While the rear wheel readings differed by only 4 kg. (9 lb), there was 25 kg. (55 lb) more weight on the left front wheel than on the right front wheel.  This difference seems excessive so we will have to have another weight-in at a later date to re-test it.

Jeff Kramer examines the 4 scale weight readings.