Experiment: Force exerted by Model S door handles on closure

[Chris TX]

If the door handle starts to retract and your fingers are in the handle, DON'T yank them out! You scrape your knuckles and fingers across the edged door surface. If you just hold your hand still, the door will sense an obstruction and pop back out. This is similar to the window or sunroof closing on your hand. If you just let it kiss your hand, it will open back up. If you pull your hand out WHILE it's touching, you're going to have a bad time.

Don't panic, the Tesla just gives little nibbles. If you show fear, it will bite.

 

[Krugerrand]

You DID say you wanted it in dummy terms.

Well, yeah. But gosh, I didn't want to feel dumb. :frown:

 

[wk057]

If the door handle starts to retract and your fingers are in the handle, DON'T yank them out! You scrape your knuckles and fingers across the edged door surface. If you just hold your hand still, the door will sense an obstruction and pop back out. This is similar to the window or sunroof closing on your hand. If you just let it kiss your hand, it will open back up. If you pull your hand out WHILE it's touching, you're going to have a bad time.

Don't panic, the Tesla just gives little nibbles. If you show fear, it will bite.

The handles don't actually pop back out. If it's locked they stay shut, and if they timed out still unlocked you have to pull still. Has been the case with each car I tried this on.

Am I missing something?

 

[Pollux]

Love the investigation. Thanks, @wk057.

 

[Chris TX]

The handles don't actually pop back out. If it's locked they stay shut, and if they timed out still unlocked you have to pull still. Has been the case with each car I tried this on.

Am I missing something?

I used to have a rear driver's side handle that would start to close right when I went to open it for one of my girls. The car was unlocked and all the door handles were presented. This one handle would just start retracting on its own, while the other ones were out. After it would close (or try to) it would open back up.

I'm guessing the original problem stems from all the doors closed and the car is in the process of locking itself (door handles and mirrors closed) then yes, the handles would attempt to close and not re-open. If that's the case, just use the other hand to hold open the handle while removing the trapped hand. Just do the thing I do with my slightly bitey cat and say "I love you, too!"

 

[wannamodels]

New to the forum and hope to order a 70D soon...I have been reading the other related thread and this one today and just had an observation...

Very entertaining thread, and somewhat useful if anyone needs to know exactly how much force is needed to open a ModelS door...as well as to not be careless with the handles!

Could it be that there is variation in push/pull-tension over all the ModelS's being discussed here? as well as any design variation? it would be great to get pics of everyone's handle action to see if there's a diff...VIN# would also be helpful for mapping.
Could it be that not everyone has the same pain-threshold? (assuming the range we are now talking about is 2-70+ based on the comments going back and forth)
Could it be that some folks have thinner/thicker skin with some being prone to small cuts?
Believe it or not the weather plays a role here too, skin is less pliable in the cold and cuts easier...

What would Nikola say? WWNT say?

 

[Chris TX]

 

[wk057]

Started off all serious sounding..... lol

 

[deonb]

@thecloud did offer to get a force gauge as well to measure the handle on his car.

Any news on that?

 

[Chris TX]

@thecloud did offer to get a force gauge as well to measure the handle on his car.

Any news on that?

Seriously. There's a spring that keeps the handle closed. When the door handle is retracting, a lever moves and allows that spring to act upon the handle. Unless someone at Tesla put a different, massive spring on there, or maybe 3-4 of them, it's not going to hurt you unless you yank your fingers away while it's closed. Only when you quickly pull your fingers away, your skin is going to scrape against the edge of the door.

I feel we're making a mountain out of a molehill. You can tell by the sarcasm in my voice on the video.

 

[wannamodels]

I feel we're making a mountain out of a molehill. You can tell by the sarcasm in my voice on the video.

Totally agree, but as one man's ceiling is another man's floor...one man's mole hill is another man's Mt Everest...Lol!


Seriously tho, thanks for the video, it does help...!

 

[deonb]

Seriously. There's a spring that keeps the handle closed. When the door handle is retracting, a lever moves and allows that spring to act upon the handle. Unless someone at Tesla put a different, massive spring on there, or maybe 3-4 of them, it's not going to hurt you unless you yank your fingers away while it's closed. Only when you quickly pull your fingers away, your skin is going to scrape against the edge of the door.

I agree with that. I also don't know how a spring can malfunction in a way that makes it tighter instead of looser. (Maybe if the coils rust together?)

But since there is a group of people contending that their door springs are tight to being up the point of being dangerous, we should be inviting them to post a measurement.

Specifically @thecloud was generously offering to obtain a force gauge to post his findings.

Discounting outright shenanigans, there's really just 2 possibilities here:

Either:
a) Some door springs are tighter than wk057's
b) Some people able to injure themselves on a door spring that's as weak as wk057's

Both options seem equally unlikely. But (a) is easier to test.

 

[cgiGuy]

The highest out of the eight was 3.87 lbs, which was ~0.2 lbs higher than the next highest. The lowest was actually 2.88 lbs, which was ~0.1 lbs lower than the next lowest. Average between the eight was about 3.36 lbs. 25% more than the highest reading would go against all of the data I have, but you're welcome to do your own testing and find data that supports that.

This is fun!

You've actually strengthened my point with additional data that you conveniently left out the first time. 2.88 to 3.87 is a 34% increase. So with all the "data that you have" which is--again--2 cars.. you've measured a 34% increase. And you have no idea if it's possible for it to be 25% or 34% more than your highest measurement on 90K other cars.

Here the edge on my car, for reference:

And here's the edge on a soda can, for reference:

So, mine are actually more like the top of the can that you mentioned.

But please tell us more about every Model S ever made...

 

[wk057]

This is fun!

You've actually strengthened my point with additional data that you conveniently left out the first time. 2.88 to 3.87 is a 34% increase. So with all the "data that you have" which is--again--2 cars.. you've measured a 34% increase. And you have no idea if it's possible for it to be 25% or 34% more than your highest measurement on 90K other cars.

Here the edge on my car, for reference:


And here's the edge on a soda can, for reference:


So, mine are actually more like the top of the can that you mentioned.

But please tell us more about every Model S ever made...

I think the force measurement data is pretty clear. Even if it is double somehow on another vehicle (not possible, IMO based on the design and evidence from testing eight handles which show a decrease in force over time as expected) it is still significantly less force than experienced in every day tasks. If someone measures a handle with 34% more force than 3.84 lbs then by all means post the data. I think I have a better chance of winning the lottery.

Also, I measured both the can and the door handle body edge with a micrometer before mentioning the can. Did you? With the micrometer still set for the measurement from the Model S (2.4mm-ish on the inside lip behind the handle, a pain to measure) it easily fits part way down onto the ridge on the bottom of the can. The ridge on the top of the can is a full mm smaller and the micrometer has a full mm of slack when placed over it while set to the door handle opening measurement.

But please tell us more about every Model S ever made...

This is far more "provocative" than my post in the other thread. In fairness I request that a mod apply negative rep for the post above also.

 

[jaguar36]

This seems like a pretty contentious thread, so I won't comment on if it can hurt or not. However testing a single sample and extrapolating those results over an entire production run is done very commonly in many industries, including big dollar industries like Aerospace. Frequently a test will be done on a single item, and that single test is used to validate that every other part, or assembly is qualified. A load factor will be applied to account for statistical variation, and if only one item is tested, it will be quite high, like 150% or even 200%.

Therefore I think its resonable to say that none of the door handles on any Tesla ever made are putting out more than 5.8lbs.

 

[Johan]

How is it not possible that wk057's measurements are all correct and from all his experience and understanding of the issue he cannot fathom how anyone could ever get hurt by a handle, but yet there is someone who has been hurt? So either you are 100% sure your conclusion is correct, and call the other person a liar, or you believe the person and integrate that knowledge in to your model and conclude that the probability of someone being hurt is no longer 0% but slightly higher.

 

[beeeerock]

All I know about this is I've enjoyed the image of wk057 wandering the house with his cool electronic fish scale, testing doors... :biggrin:

And that's because it's something I can admit I'd probably do too, with a new toy like that... and i know my wife would give me no end of grief about me wasting my time doing it! She'd probably enjoy telling our friends about what an idiot I was too... much like the time I got a deal on an IR temperature gun at Costco... LOL

wk057, we're probably related somehow. :biggrin:

 

[wk057]

All I know about this is I've enjoyed the image of wk057 wandering the house with his cool electronic fish scale, testing doors... :biggrin:

And that's because it's something I can admit I'd probably do too, with a new toy like that... and i know my wife would give me no end of grief about me wasting my time doing it! She'd probably enjoy telling our friends about what an idiot I was too... much like the time I got a deal on an IR temperature gun at Costco... LOL

wk057, we're probably related somehow. :biggrin:

^ This.

 

[bonnie]

This seems like a pretty contentious thread, so I won't comment on if it can hurt or not. However testing a single sample and extrapolating those results over an entire production run is done very commonly in many industries, including big dollar industries like Aerospace. Frequently a test will be done on a single item, and that single test is used to validate that every other part, or assembly is qualified. A load factor will be applied to account for statistical variation, and if only one item is tested, it will be quite high, like 150% or even 200%.

Therefore I think its resonable to say that none of the door handles on any Tesla ever made are putting out more than 5.8lbs.

The industries I'm intimately familiar with (aerospace, medical device, chip manufacture, pharma, biologics) require a First Article sample for the first part received from a vendor, with a sample size of 1, to ensure that the vendor is capable of meeting print. But that's only First Article. With every single shipment, sampling is done based on size of shipment, past vendor history, criticality of component on device performance, and other factors dependent upon industry and usage model. I'm most familiar with C=0 acceptance testing, but MIL-SPEC-105, for instance, will set a minimal acceptable quality level (AQL) for a sample size.

First Article (in the industries where I've worked) checks every single dimension against the print. Sampling plans typically select key part attributes (not everything), with the intention that if one of those measurements has drifted, then other dimensions will likely be out of spec also. Usually the selected dimensions are critical to the proper performance of the part. Do parts slip through with sampling? Heck yes. That's why there is also in-process inspections for most devices.

First Article is not equivalent to acceptance testing. It is part of verification and qualification of the vendor. You only repeat a First Article when either the part revs (Like from Rev A to Rev B) or you have a new vendor for an existing part. I know of no industry that would accept parts for production based on only sampling one.

 

[thecloud]

@thecloud did offer to get a force gauge as well to measure the handle on his car.

Any news on that?

An attempted delivery was made this morning but I missed it. Should hopefully have it tomorrow.