Cadillac Spindle Failure

Find answers to all of your automotive questions here
Post Reply
Posts: 156
Joined: Tue Jul 24, 2012 6:15 pm
Location: Morristown, NJ

Cadillac Spindle Failure

Post by oldcarfudd » Fri Jul 29, 2016 6:59 pm

My wife, Joan, and I had a material failure with our single-cylinder Cadillac on a tour last week. The right front spindle broke, the wheel came off, the spindle dug into the unpaved shoulder, the car slued abruptly to the left, and Joan was thrown out onto the pavement and landed on the back of her head.
The disabled Cadillac after the spindle failure.
photo.JPG (91.14 KiB) Viewed 7608 times
The broken spindle was dragged through the unpaved shoulder, sluing the car.
photo-1.JPG (125.18 KiB) Viewed 7608 times
This end of the stud broke off, letting the wheel come off.
Good news first: Joan had nothing broken, but she had a concussion and was treated at the local hospital. She’s had several days of severe but (fortunately) diminishing headaches. She’s on the mend, and expects to resume riding horses – gently at first! – in a few days. The failure took place on the Canadian HCCA tour; we’re grateful to the passersby, firefighters, police, ambulance squad, trouble-truck drivers, and other tourists who took great care of us and our broken car.

Now let’s talk about the car. Cadillac front wheels were originally held on with a three-fingered arrangement secured with an expanding nut and bolt. It worked well, but the fingers tend to fatigue after a century of disassembly and reassembly and other owner abuse. They can break. There was a serious accident on a tour in Kansas in 1997. As a result, Don Rising published an article in the Early Cadillac Bulletin. It described a fix consisting of drilling and tapping the end of the spindle for a Grade 8 stud, pinned or Loctited in place to keep it from unscrewing itself, and using a conventional castellated and cottered nut to hold the wheel. Functionally, it turned the Cadillac spindle into a Model T Ford spindle.

My car had had this repair, with one exception. The stud wasn’t pinned or Loctited; it was spot-welded. The stud broke just inside the end of the drilled and tapped spindle, where no visual inspection could have discovered a developing crack before it failed.

I called Don Rising, seeking advice. He recommended I speak with Garyl Turley. I did, and I think I learned a lot. Garyl said that a weld in a hard steel like a Grade 8 bolt creates a metallurgical notch – that is, a stress point – that locally changes the strength of the material. Also, a Grade 8 stud is very high in tensile strength – you can crank a nut down on it really, really tight – but it’s brittle - it doesn’t want to bend very much if it’s subjected to up-and-down loads, like a bouncing wheel. He suggested three alternative repairs, in ascending order of complexity and cost:

1. Redrill and retap the spindle and make a new stud. But use a chrome-carbon alloy steel like 4130, which has plenty of tensile strength but isn’t as brittle as a Grade 8 stud. And, rather than welding or brazing it, hold it in place with industrial strength Loctite. This stuff is used in aircraft construction. Wrenches won’t take it apart, so it won’t unscrew itself. If you ever need to take it apart, the heat of a propane torch will loosen the Loctite without affecting the metallurgy of either the stud or the spindle.

2. Cut off the horizontal part of the spindle – the part the wheel revolves around – leaving the vertical part that the king pin goes through and the arm that goes to the tie rod. Machine a new horizontal part, including an integral threaded end to hold the wheel on, out of a modern mild steel like 1018 that’s metallurgically compatible with the mild steel used on the original part. Deep weld the parts together, and x-ray or ultrasound the finished part to be sure the weld is thorough. Metallurgic compatibility is useless if there’s no metal in the middle.

3. Machine a whole new spindle, in three parts – the horizontal part to hold the wheel, the vertical part to hold the king pin, and the right-angle part to hold the tie rod – and deep weld them together. X-ray or ultrasound the welds as in the prior paragraph. If this route is chosen, make the whole thing out of 4130 – better metal all around than the original.

Garyl further suggested that this would be an excellent discussion topic at the Single Cylinder Cadillac Register meeting at the Hershey tent in October.

I don’t profess to know a damn thing about metallurgy. The numbers of these steels might as well be sections of the Internal Revenue Code. If anyone out there has anything to add or any other ideas to pass on, please do so. If anyone wants to send this thread on to Facebook, or to a region newsletter, or to another car club, or to the Gazette (although John Meyer doesn’t like picking up stuff from the website), please feel free. My purpose in posting this is to start a conversation, and possibly to spare someone else an accident. My little Cadillac has been my favorite car for several years, but I will not drive it again until one of these three repairs, or an alternative suggested by someone knowledgeable, has been done. This accident happened on a dead flat road with no traffic. It could have been very much worse. We were very lucky; I don’t intend to push my luck.

Gil Fitzhugh the Elder

Posts: 116
Joined: Sun Aug 30, 2009 11:39 am
Location: Grass Valley, CA

Re: Cadillac Spindle Failure

Post by FullerMetz » Tue Aug 02, 2016 10:23 pm

Ooh, Ouch!
Glad you are okay and that your wife will be.
Losing a front wheel that way is a scary thing, almost as scary as being thrown by a horse. Tell your wife to be careful there also. My wife broke two ribs that way some years back. She still has pain because of it.

Metallurgy is a specialty requiring a lot of areas of expertise. I know enough about it to know that I know very little about it. I do have enough engineering and welding background to know that I do not like welded spindles Garyl Turley is probably right about the metal numbers being stronger than the original spindles were (I met him some years ago). And definitely, X-ray by a specialist at least would have to be done to verify good enough penetration.

I would consider choice number one myself, but only after I could closely examine the existing spindle myself to see if enough good and usable material is present to maintain strength and integrity. Hollowing out spindles doesn't thrill me exactly.
Another option I would suggest. I know it is considered blasphemy by some in the Horseless Carriage world. And I do not have an appropriate Cadillac nearby to look at to see if it is a viable idea or not. But I have seen a couple cases of model T Ford spindles being modified to work in non-Ford front axles. Most smaller Horseless Carriages have smaller enough spindles that a Ford spindle can be cut back or shortened a little and made to work by insetting the steering arm and having it properly welded. The advantage to doing this, is that you begin with the single piece for both vertical and horizontal axis of the spindle. That covers the area of greatest safety concern. By insetting the arm, the use stresses are distributed into both the inside and outside of the weld (X-ray still advisable). Some adaption of the wheel bearings and hubs may be necessary. Minor adaption of king pin or bushings should be simple to do.

Again, I am so pleased that both of you are okay. The car can be fixed for sure. It could be replaced if worse had happened. Loved ones and ourselves? Much more important!
Drive carefully, and enjoy!

Posts: 6
Joined: Mon Oct 12, 2015 1:37 pm

Re: Cadillac Spindle Failure

Post by nsbrassnut » Sat Aug 06, 2016 5:48 pm

Hi Gill

Glad to hear that you and your wife are OK after the Cadillac accident.

I would suggest that you also bring you post to the Early Cadillac Group (Yahoo) where I expect it would trigger a good discussion.

In the meantime, when I was restoring my ’05 Cadillac I gave some thought to the front wheel lock issue. I have read of the earlier accident and wanted to look at options to improve the bearing retention lock.

I decided that I didn’t like the style of modification that was used on your Cadillac. It’s not really as good as some think it is. Drilling out the original spindle and re-threading it for a bolt can weaken the spindle as this reduces the remaining wall thickness that holds the wheel.

Also installed a plane threaded rod with only a pin for retention in the spindle and a single castle nut with cotter pin is no better and potentially worse than the original design. In engineering terms, the original design and the modification that appears in your picture are “single point of failure” designs. Meaning that only one part has to fail for the wheel to come off. This concern also applies to the Cadillac brakes which have no secondary parking/emergency brake either. Having experienced brake failure in one of these Cadillacs, I can say it is not fun (remember Maine 2014).

For the original, if the fingers bread off the retaining bolt, then the bolt can unwind and the wheel come off. In the bolt medication approach, if the pin fails, the bolt can unwind or if the cotter pin fails then the castle nut can unwind.

If you look again at the Ford T wheel bearing system you can see that it designed with a “two point of failure” design where two things have to fail for the wheel to come off. The T spindle also has a keyway which is part of the retention system. The T front wheel race is threaded, then there is an anti-rotation keyed washer and then the castle nut. The castle nut is tightened against the bearing race. This jams the two together like a lock nut and both. The keyed washer helps prevent the two from turning helping to kept them locked. Then there is the cotter pin on the end as a back up to prevent the castle nut from coming off. So for the T wheel to come off, the bearing and castle nuts have to work loose and the cotter pin has to break, so two points of failure before it comes apart.

The modified Cadillac spindles that I have seen pictures of are not as protected as the T. the Cadillac hubcap is short and there is not much space to have double lock nuts on the bolt plus a cotter pin to provide the double protection.

I came up with an alternate double lock method for my Cadillac that requires only a small amount of modification to the original parts to provide a much better retainer lock. I drilled and threaded for a set screw (1/8”) that is in line with the end of the spindle rim. Then ground a matching recess in the end of the spindle. This recess is rounded to prevent creating a potential stress riser in the spindle. Now the when the retainer is threaded into the spindle and the bearing is set, the end is rotated so that the set screw aligns with the recess and gently tightened (too much force could bend the retainer). Then tighten the set screw into the recess which now prevents the retainer from turning out of the spindle. Then a washer is placed under the head of the original locking bolt covers the set screw and the original locking bolt is tightened normally spreading the end of the retainer inside the spindle just like before. Now for the wheel to come off both the inside “fingers” have to break and the set screw would have to break converting this to a two point of failure locking mechanism.

I also did some comparing of a T spindle and a Cadillac spindle. The height of the T spindle is similar and it will fit in the end of the Cadillac axle. However the length of the spindle is different and diameter where the bearings slide on. Later style T spindles with the bent steering arm would need to be used to for the tie rod to clear the radius rods. Also, the steering arm is shorter than he Cadillac which would make the steering more sensitive than it already is.

You could always consider installing another set of original Cadillac spindles and use a modified original bearing retainer.

Good luck with your Cadillac repairs.

From the Great White North - Canada

Posts: 156
Joined: Tue Jul 24, 2012 6:15 pm
Location: Morristown, NJ

Re: Cadillac Spindle Failure

Post by oldcarfudd » Tue Aug 09, 2016 4:10 pm

Jeff, thank you for the additional suggestions. I agree others will want to hear about them. I e-mailed Steve Hammett, who runs the Early Cadillac Yahoo group, suggesting he repost what I had written, or at least link to it, but I haven't heard back from Steve. I got into some kind of computer glitch with my Yahoo account a year or so ago and have never been able to resolve it; otherwise, I'd have posted this stuff there myself (as well as here). Would you please post it to that forum, or post a link to this thread? Many thanks.


Posts: 6
Joined: Mon Oct 12, 2015 1:37 pm

Re: Cadillac Spindle Failure

Post by nsbrassnut » Sun Sep 04, 2016 4:43 pm

Hi Gill

I finally got your notes posted on the Early Cadillac group forum as well. I have been surprised by the small amount of follow up it has gathered so far.

Drive Safe

Nova Scotia

Posts: 3
Joined: Sun Sep 25, 2016 12:54 pm

Re: Cadillac Spindle Failure

Post by 1cylnut » Mon Sep 26, 2016 3:04 pm

Hi Gil,
So glad you both survived the ordeal with minimum damage. I am out of Cadillac at the present so I wont go into design but Garyl's advice seems good. I did have a spindle failure on my CDO and repaired it by Garyl's #2 method but was able to press the new heat treated 4130 shaft through the body of the spindle and make the weld on the inside so it was not in a "stressed" area. Perhaps that might be an option here. For the #1 method here, I would have preferred fine thread. If you could go one size larger and still have enough meat in the spindle, then use a grade 5 bolt with Loctite #680 I think you would be fine - defintiely don't tack weld. Be careful with using 4130 - heat treat can be critical - best to buy the material in a heat treated condition rather that just annealed. The heat treated version machines fine and is a good balance between toughness and hardness.
See you at Hershey,
Bill Ottemann

Posts: 7
Joined: Wed Nov 23, 2016 2:36 am

Re: Cadillac Spindle Failure

Post by Johnwick88 » Tue Dec 20, 2016 11:52 pm

Sincerely sorry But I think she would have gone back on the same ride.


Post Reply