Forum

Meld dig ind i fællesskabet
Meld dig indLog ind

Check Your Nuts! Found Assembly Error During Hitch Install

Today I undertook the installation of a Torklift trailer hitch on my new Model S and encountered a scary situation. The Torklift hitch is a custom designed hitch for the Model S and comes with detailed instructions on how to remove the rear bodywork, bottom panels, and bumper crossbars to install the hitch receiver. It installs without any drilling, using existing bolts and nus that are part of the Tesla bumper system. It is a great piece and will allow us to use a bicycle carrier with nothing touching the car.

Installation is quite complex, and for most people would be a job for a body shop. However, we have a lift in our barn and all the needed tools, so I tackled it myself. All went well until I removed the bumper cover. Immediately inside the bumper cover is a second stamped steel bumper cross member that is attached with eight nuts to studs that protrude from the aluminum body. This steel cross member is actually a cover for an extruded aluminum cross member that is the main structural element inside the rear bumper. The instructions are to remove the eight nuts holdng the steel cross member with a 17mm socket wrench, then remove eight more nuts that hold on the aluminum crossbar, using a 15mm socket wrench. The trailer hitch is then mounted behind (forward on the car) both crossbars using the original nuts to hold all three crosspieces (including the new hitch) in place.

Imagine my shock when I saw the following pictures (I hope these links work this is my first try at embedding links on the forum). When I pulled the plastic bumper cover away, it was immediately obvious that Tesla had at least one serious quality control issue when assembling my car. Three of the four nuts holding the right side of the steel bumper crossmember were barely threaded onto their respective studs. They were not even hand tight, but very obviously just started onto the studs and far away fom being seated. The fourth nut was properly torqued to 40 ft. lbs. (the torque recommended in the hitch instructions).

If I had not seen this myself I might not have believed it. It was obvious, and in my opinion, the bumper cover installer would have had difficulty missing the loose nuts left behind by the cross member installer. I do not know if that function is robotic, but doubt it. It is serendipitous that these loose nuts are three of the very fitments that I was going to be removing anyway to position the hitch assembly.

This is disturbing on several fronts. How many people with ths condition would uncover it? The loose nuts are fully enclosed in the bumper cavity and cannot be seen. I presume that only a small fraction of Model S owners will install hitches. It is a big job, and requires removing dozens of screws, the entire rear body panel system, carpets, hatch seal, interior panels, rear key sensor etc. Very few people will ever see these nuts in their cars, and they cannot be inpected without hours of surgery to get the external panels out of the way. These bolts are part of the rear crash protection system and are critical to safety. I presume that this means that our car did not meet its safety design specifications, and we never would have known if it were not for my home wrenching project.

Lastly, if assembly quality is this haphazard, I wonder what else isn't bolted tight where it cannot be seen? Part of me wants to ask TM to do a thorough build-level inspection, removing all interior and exterior panels to ensure everything is properly tightened and torqued. Perhaps this is an isolated assemby error, but I worry that it means smething more fundamental - that in the rush to meet its production objectives, TM is having quality issues on the assembly line. I love my car, but it isn't going to be quite the same after this experience.

Links to detailed pictures follow.

https://www.dropbox.com/s/oplyslg0si7axzi/2013-02-14%2016.03.02.jpg

https://www.dropbox.com/s/rhmh0vrh51w544m/2013-02-14%2016.03.48.jpg

https://www.dropbox.com/s/j0zrkpcauw52n78/2013-02-14%2018.12.58.jpg

https://www.dropbox.com/s/w1ivg4x5b81dawp/2013-02-14%2018.12.43.jpg

https://www.dropbox.com/s/cqxx8xxs5j2psmz/2013-02-14%2018.11.52.jpg

https://www.dropbox.com/s/yg2uyl1biihr7q8/2013-02-14%2018.11.59.jpg

https://www.dropbox.com/s/7qctthfogh7flgu/2013-02-14%2018.12.14.jpg

Yikes! That is troubling indeed. I hope you've notified someone at Tesla about this. I would feel the same about wanting them to tear down the entire car to make sure they didn't miss any others.

Thanks for sharing.

This is not good. Share with Tesla. Email as high up in the chain (georgeB for example) as you can.

I agree with RedShift. Email as high as you can, this requires inspection and perhaps even recall for some cars to fix it if there are many that have this issue. I hope it is isolated error.

@Pungoteague_Dave - Absolutely you should email George Blankenship about this if you have not already. Please post his reply. Thx.

Sorry to hear about your loose nuts. I was considering installing a Torklift hitch also - this will speed up my decision. Do you have any more installation pictures? I would like to see how the receiver hitch is attached. It appears that they either cut a hole or remove the corresponding trim piece under the car. The Torklift web site says that no cutting is needed for the installation.

That probably should have been the first step, not the second.

I have a Model S on order and this lack of quality control is disturbing, especially in a component that has to do with safety (the bumper) and where the defect would not be spotted in any normal operation until the component was stressed (as in an accident). Could you please forward this information to as many people within the Tesla organization as necessary and as high up as necessary to let them know what you found so that they can correct the fundamental processes that led to this failure in their assembly system. As a new car company Tesla needs a strong manufacturing quality control system in place and the person or team who put on this component in the factory needs to know that their level of workmanship must improve. Tesla will not survive long term if they get a reputation for shoddy workmanship and that would not benefit either us or them. It would be a great help if you could keep us informed as to the response that you get. Thanks for the excellent pictures.

May I suggest that you make this thread "private" do only owners and reservation holders can see it?

And yes, George Blankenship needs to see these pictures. Scary!

Wow that's bad...

Did you find any proof of the parking-sensor cable tree being there (preinstalled as some suggested)...?

Hmm, would add washers and lockers.

I will definitely send this information to TM management.

I plan to set up a thread on the hitch install with links to more pictures. There is a video on the manufacturer's website. I did have a few fit issues that required some minor grinding on the hitch. Installation does require cutting a pretty big hole in the bottom rear body cover, including the forward part of the plastic chrome finisher at the bottom of the bumper. They provide a template and suggest using a 4" hole saw on both ends of the oblong hole to get it started. However, I did not have a hole saw that big, so just masked the surface to protect the paint and chrome, and transferred the template to the tape, drilled a starter hole, and used a jig saw to do the job.

Cutting the bumper cover was the scariest part of the installation and I suppose could marginally affect airflow and therefore range, but I expect not by much. When the bike carrier and two road bikes are on the car, it may have a measurable range impact, as it is kind of like pulling a parachute behind a vehicle designed for least air resistance. I decided to do it because we can't use regular strap-on bike carriers with the aluminum body, and since the car is designed with specific attachment points for a roof rack, Tesla must have testing saying exterior wind disruption isn't terribly bad to range. I figure that a rear hitch mounted rack is less disruptive than having two bikes mounted on the roof.

I posted this just after getting the hitch on the car late last night, and still have to button up the installation and reinstall the wheels, will contact TM later.

It was surprising to see that there are no lock washers or locking nuts on the bumper installation, but I am certain these did not loosen on their own. They were too far backed off the bumper plate surface, and there was no paint disturbance indicating they had ever been seated. The other five nuts that were seated properly did mark the black paint on the steel bumper crossbeam, so it is pretty clear what occurred. Someone just forgot or missed these nuts after they were hand-started onto the fitment studs. The one nut on the right side that was tight, was enough to keep the assembly from rattling, and there is enough friction on the painted studs that the three loose nuts did not fall off in normal driving. I wonder if TM can figure out who did the bumper installation based on VIN, perhaps for a little "reeducation"?

Digging into the car like this is a real education. I am more impressed than ever with the engineering that went not these cars. While not up to Mercedes or BMW standards for everything (for example, they use regular wire ties or plastic tape where the Germans would have special little attachment fittings, and some materials are pretty cheap, such as the carpets and interior LED cabin light fixtures), but the mechanicals, usually hidden behind the body panels, are amazing. The compact motor looks like it would be at home on a space ship, and the suspension bits are worthy of a Formula racing car. It s clear that the money went where it counts. But it is also clear that TM is still working on getting its act together as a manufacturer.

I have checked my nuts now, both are ok. ;-)

The pictures of the hardly-threaded nuts has all the markings of "got interrupted part way through the job, forgot I hadn't tightened them all, someone else put the bumper on thinking everything was finished."

This does not look like a real safety issue to me. The bumper does not absorb much energy in a rear end collision and would not detach anyway since it would be pushed in. It would be more likely to fly off from a sideswipe or rear lateral hit, but neither of these would threaten occupants.

That said, it's a clear manufacturing error, and as such it carries the implication that their process for verifying work is (or was at the time) broken at least in that part of the assembly line.

I think Tesla should inspect all rear and front bumper attachments ASAP and obviously verify that assembly process.

I don't think this has broader implications for structural integrity, because the structure and safety cage are welded by robots. The battery is bolted on but that part is so heavy it would just fall out if improperly attached.

Lots of bolts in suspension assembly and so forth. That's a different part of the line and the shakedown would catch it, but I think Tesla should inspect anyway. Especially since safety would definitely be compromised by a poorly attached suspension piece, while inspection is simple due to easy access with the car raised and wheels off. It's thus a high payoff, low cost inspection.

Sampling inspections could also be done on seat attachments and pedal assembly.

Ultimately Tesla could just X-Ray the whole car as they come out the assembly line. The images could be subject to computer analysis in an automated process. No idea what the unit cost would be but car sized machines are routinely used in customs now.

Instead of X-ray, you just do QA inspections as the car is being built. Much easier. Dave, what did Tesla say when you called them? I would think they would do an in field inspection of the cars before and after this car was built to see if this was an isolated incident or not.

Makes me recall the NGeo movie of the assemblers trying to beat the stopwatch ...

jeroens,

My dissassembly did confirm that there is no wiring at the rear of the car for parking assist to be added later. I had the entire rear lower body and bumper system off the car and inspected closely. The only wires encountered were connected to interior and exterior lights, to the motor and brakes, wheel sensors, a rear sensing ring, and the battery system. The hitch instructions had me relocate the sensing ring fom the body sheet metal to the rear of the hitch frame so it remains directly behind the plastic bumper cover. I presume this is a key sensor, but am not sure of that. Several TM employees, including two at delivery in Rockville confirmed that the car has no hardware provision for parking sensors.

danielccc,

I do think that structural integrity would be compromised in an accident without those nuts being properly torqued. In one of the review videos Elon can be seen describing the safety afforded by the "massive" rear cross members that are there to protect rear facing occupants in the jump seats. It looks to me that not having three of the four nuts on one side of a body member that is there to deflect the force of a rear collision would substantially affect structural integrity and safety. The mounting studs are welded into the car's aluminum body structure, but the two cross members are stamped steel over a heavy aluminum extrusion. They look like they are engineered specifically as the primary rear crash protection element that both ties everything together and is the "crash bar" in the occupant protection system. I am no engineer, but missing a large portion of its attachment points (three of four fitment nuts on the right side) would have to cause a measurable loss of crash worthiness.

P_Dave, I am an engineer, and I'll note that the bolts are going through the holes. So, in the event of a rear end collision, where does the bar go? It will not detach. It will pull on the bolts and lock into place as it does so.

Sort of.

The nuts transmit force through their contact surface area, by friction. Without them, all the force goes to the edge of each hole. These can deform. The bar can shift more than if the nuts were properly torqued. Also, the joint can rotate more freely.

I had forgotten about the jump seats. You are right that, in some geometries of impact, the lack of proper attachment could lead to greater displacement than otherwise near the bumper. But the jump seats are quite far from the area because the car is so long.

I am not trivializing this but I don't see a major safety issue. I am more concerned about what it says of the assembly process. It might be a one-off, but as I said, Tesla should inspect front and rear bumper bolts as well as suspension assembly, plus some other bolted assemblies on a sampling (statistical) basis.

danielccc,

The jump seats are NOT far from the bumper as you assert. A kid's knees are less than five inches in front of these structures. The footwell is immediately in front of the rear body panel to which the bumper reinforcing cross members attach, and a person sitting there facing back would rest their feet against the back of the rear body panel and bumper structure. There is zero air space between the footwell and the back of the car.

Dave, what did Tesla say when you called them?

@shop, the thing that attracts me about X-ray inspections is that Tesla could keep the X-ray images for the life of the car.

So this means is that they could verify that problems like this were not introduced after it left the factory, for example by a careless mechanic (side note, I brought a Chevy home from the mechanic over ten years ago and a part of the front assembly was missing every bolt on the top side after some work on the radiator).

Another thing is that, should a mistake like this one be discovered, it is easy to review the images of all the cars that went out the door and update the image analysis software to catch the problem from that point forward.

Look at this image to see just how useful this could be.

@P_Dave, If there is zero space then there isn't much protection there regardless. Honestly I have not given them a lot of thought because I am not interested in them.

They will be the safest place in the car in front and side collisions, but rear ones? Pretty marginal I would think in the best of cases, especially if there is an "interface mismatch", meaning you get hit by an tall SUV or truck.

Do you have jump seats?

PD;
daniel's main point seems to me to be that in an impact, the bar is being pressed, not pulled. Nuts only prevent pulling. Almost impossible to imagine how that could happen.

Brian,
No really, if the impact to the bumper is not uniform when it is deform from the impact it will bow out everywhere those bolts are not applied fully. I think I will purchase my Torklift trailer hitch sooner than later.

See the engineering analysis above. The boron steel can deform, but that degree of impact is enough to drive the car forward into its proper crumple zone (frunk).

This particular set of nuts may not be a big issue, but that put doubts in my mind as to where else the nuts are not tight or even missing.
They are trying to beat the clock, so issues like that are almost inevitable.
Hopefully it was a one-off and not a regular occurrence.

The response from Tesla management is going to be critical. PLEASE post it immediately on receipt. I agree, this is a biggie. While it does sound as though the worker concerned was interrupted and the line moved on, we are all human and, failing proper checks and balances, will happen again.

Sorry, was in a faculty meeting all day so did not yet contact TM. Will do first thing Monday and report back.

The hitch worked great but does reduce range. We drove 170 miles with it and a bike on a hitch rack last night. We started with full max range charge (272 rated at start) and drove normally. Mostly 65 mph, some 75. We had 10 miles left on arrival, so lost about 90 miles of rated range to temperature, slightly above average speed, and the rear rack. The hitch/rack/bike combo seemed to affect range about 10-15%, as we averaged 404 kwh, and normally average 340-350 on that trip at those temps (about 40 degrees). However, we did have the headlights on as it was dark the entire way, and all of our prior trips have been with the lights off during the day.

Since we had almost no cushion with a max range charge for this trip, I am further reducing my range expectation for this car. I now think of the 85 kwh S as a 175 mile range car for normal trip planning, but 155-160 with the bike on the back, and less as the battery ages... Not what we had hoped for when we signed up. But we still love the car, just can't use it for some of our commutes to/from the city.

danielccc, I don't have the jump seats and won't as my kids are all in their 20's.


X Deutschland Site Besuchen