Welcome to Tesla Motors Club
Discuss Tesla's Model S, Model 3, Model X, Model Y, Cybertruck, Roadster and More.
Register

Model Y road noise/rumble primary cause

This site may earn commission on affiliate links.
One of the root causes: relatively weak static and dynamic torsional rigidity. This problem is not specifically for MY, it also applies to all current Tesla models more or less. Unless Tesla completely revised the design from the ground up, this architecture-level issue cannot be solved simply by "patching here and there".

Looking at any Tesla model's frame structure, we can see there is no effective structural component to resist the torsional stress generated from relative twisting between the front and rear bulkhead. For Model 3 and Y, it is the battery pack and also the glass roof acting as the stress component to counter the twisting force.

Tesla-Model-Y-vs-Model-3-underbody.jpg


There are only horizontal beams inside the battery pack/floorpan, and the pack is thin (compared to the total height of the car body), it almost can be treated as a 2D plane. Therefore a lot of the torsional stress is applied to M3/Y's glass roof. Glass is never good at resisting this type of stress, this is the reason why we see so many glass roof crack complaints. All these factors lead to Model 3/Y'sv sub-optimal static torsional rigidity.

Generally speaking, static torsional rigidity is positively correlated to dynamic torsional rigidity. A vehicle's dynamic torsional rigidity is measured as its 1st-order resonant frequency along the diagonal direction of its frame. Low dynamic rigidity means the vehicle body will resonant at a lower frequency.

No road surface is 100% flat. Small, and countless imperfections of the road surface continuously generate pulsing excitations. Once the vehicle picks up speed, the dominant rhythm of the pulsing excitation reaches near the car body's 1st-order resonant frequency, the vibration will be dramatically amplified, then you will start to hear those annoying rumbling and humming noise.

Some side notes:

1. Traditional ICE cars (especially RWD layout) do not have this issue: the transmission tunnel built into the body frame acts as an effective and sturdy "torque tube". Some fancy sports cars even use a torque tube or torque beam explicitly to further improve the rigidity, for example:

- Lexus LFA (to flex LFA's body, one will need to "twist" that gigantic black tube, which takes enormous effort and is almost impossible under all considerable scenarios):

LexusLFATorqueCube.jpg


- Mazda Miata (see the torsion beam which connects the transmission and the rear differential)

546b39a62f36a_-_p1j07859l-lg.jpg


2. Some EVs also build a "tunnel" to connect their front and rear bulkhead to guarantee torsional rigidity, for example, Porsche Taycan:

Porsche-Taycan-30.png


3. Many luxury cars have sophisticated torsional components/ribs built into the floorpan and welded with the transmission tunnel, this is also a very effective technique to strengthen torsional rigidity. Take a look at Genesis G90's bowl-shaped components between the horizontal and lateral ribs in its floorpan, and how they are welded to the transmission tunnel.

6-hero-2-safety-without-compromise.jpg


4. There is a reason why flagship luxury sedans/SUVs usually have high rigidity and long-wheelbase: high rigidity => high resonant frequency; long-wheelbase => higher vehicle speed for the road surface impact forces to reach the same vibration frequency, compared to the shorter wheelbase. High car body resonant frequency + higher speed to receive excitation at that frequency => the vehicle will only resonant at a speed which is far higher than its top speed, so many latest car models (especially luxury cars) seldom resonant (very quiet) under all reachable speed.

It is hard for EV to use the above structural reinforcement techniques because it needs to host lots of battery cells. Although the Taycan uses a center tunnel to boost rigidity, it sacrifices space efficiency and battery placement/capacity/range, which happens to be Tesla's selling points. For M3 and MY, weak torsional rigidity + relatively short wheelbase is a perfect recipe for the car body to resonant starting at low speed, and the 2nd-order/3rd-order...resonant frequencies will kick in one-by-one as the vehicle speed increases, so you will always hear the annoying rumbling during most of your driving time.

Car body/frame structure design is in fact much harder than designing the powerplant/drivetrain, it requires lots of know-how and experiences, cannot be learned and acquired in a fast-food style, and no silver bullet exists most of the time, which makes it a fundamental competitive/differentiating factor in the industry. With that being said, it is not uncommon to see automakers outsourcing engines, transmissions, software; but you rarely see any major automaker outsourcing its body/structure design.
 
As I get ready to replace my 10 year old Leaf, my first choice is the Audi Q4-e-tron, but if that does not meet my needs, I am still considering a MY. As an engineer, structural rigidity is a big deal to me. I plan to compare the Q4 to the MY is this respect as best I can by the seat-of-my pants. I have a Boxster S for occasional use and it has an extremely rigid structure which helps provide it's fine ride and handling. I have heard that the MY is a bit lacking in the structural area and this original post does a great job of explaining why that is. Also, rumbling/road noise can be reduced with 600lbs of sound deadening materials which is what many luxury cars have, but that would negatively affect range in an EV. The best sound deadening materials are dense like lead = heavy. It does no good to cram foam into open areas to reduce noise.
 
Thank you very much for this. Given the fact that this issue seems to plague similarly designed EVs (Audi E-Tron, Mercedes EQC, Mustang Mach-E to lesser extent) this is clearly as yet an unresolved problem with modern EV SUV design. Ultimately very frustrating but also very enlightening. I imagine ironically any non-purpose-built EVs may actually excel in this area because of their ICE-compatible designs (thinking Volvo XC40 Recharge, Kia Niro EV, Hyundai Kona EV, etc).

The main thing I appreciate is that you've just saved me all manner of trouble and grief trying to "fix" this fundamental design flaw of the Model Y with band-aids like Dynamat and wasted money on new tires and shocks. Seems like it's "cut my losses" time...
 
  • Like
Reactions: polyphonic54
This "analysis" sounds like complete BS. Here's the proof that it's not worth reading at all: "this is the reason why we see so many glass roof crack complaints"

One of the root causes: relatively weak static and dynamic torsional rigidity. This problem is not specifically for MY, it also applies to all current Tesla models more or less. Unless Tesla completely revised the design from the ground up, this architecture-level issue cannot be solved simply by "patching here and there".

Looking at any Tesla model's frame structure, we can see there is no effective structural component to resist the torsional stress generated from relative twisting between the front and rear bulkhead. For Model 3 and Y, it is the battery pack and also the glass roof acting as the stress component to counter the twisting force.

View attachment 662770

There are only horizontal beams inside the battery pack/floorpan, and the pack is thin (compared to the total height of the car body), it almost can be treated as a 2D plane. Therefore a lot of the torsional stress is applied to M3/Y's glass roof. Glass is never good at resisting this type of stress, this is the reason why we see so many glass roof crack complaints. All these factors lead to Model 3/Y'sv sub-optimal static torsional rigidity.

Generally speaking, static torsional rigidity is positively correlated to dynamic torsional rigidity. A vehicle's dynamic torsional rigidity is measured as its 1st-order resonant frequency along the diagonal direction of its frame. Low dynamic rigidity means the vehicle body will resonant at a lower frequency.

No road surface is 100% flat. Small, and countless imperfections of the road surface continuously generate pulsing excitations. Once the vehicle picks up speed, the dominant rhythm of the pulsing excitation reaches near the car body's 1st-order resonant frequency, the vibration will be dramatically amplified, then you will start to hear those annoying rumbling and humming noise.

Some side notes:

1. Traditional ICE cars (especially RWD layout) do not have this issue: the transmission tunnel built into the body frame acts as an effective and sturdy "torque tube". Some fancy sports cars even use a torque tube or torque beam explicitly to further improve the rigidity, for example:

- Lexus LFA (to flex LFA's body, one will need to "twist" that gigantic black tube, which takes enormous effort and is almost impossible under all considerable scenarios):

View attachment 662764

- Mazda Miata (see the torsion beam which connects the transmission and the rear differential)

View attachment 662765

2. Some EVs also build a "tunnel" to connect their front and rear bulkhead to guarantee torsional rigidity, for example, Porsche Taycan:

View attachment 662772

3. Many luxury cars have sophisticated torsional components/ribs built into the floorpan and welded with the transmission tunnel, this is also a very effective technique to strengthen torsional rigidity. Take a look at Genesis G90's bowl-shaped components between the horizontal and lateral ribs in its floorpan, and how they are welded to the transmission tunnel.

View attachment 662766

4. There is a reason why flagship luxury sedans/SUVs usually have high rigidity and long-wheelbase: high rigidity => high resonant frequency; long-wheelbase => higher vehicle speed for the road surface impact forces to reach the same vibration frequency, compared to the shorter wheelbase. High car body resonant frequency + higher speed to receive excitation at that frequency => the vehicle will only resonant at a speed which is far higher than its top speed, so many latest car models (especially luxury cars) seldom resonant (very quiet) under all reachable speed.

It is hard for EV to use the above structural reinforcement techniques because it needs to host lots of battery cells. Although the Taycan uses a center tunnel to boost rigidity, it sacrifices space efficiency and battery placement/capacity/range, which happens to be Tesla's selling points. For M3 and MY, weak torsional rigidity + relatively short wheelbase is a perfect recipe for the car body to resonant starting at low speed, and the 2nd-order/3rd-order...resonant frequencies will kick in one-by-one as the vehicle speed increases, so you will always hear the annoying rumbling during most of your driving time.

Car body/frame structure design is in fact much harder than designing the powerplant/drivetrain, it requires lots of know-how and experiences, cannot be learned and acquired in a fast-food style, and no silver bullet exists most of the time, which makes it a fundamental competitive/differentiating factor in the industry. With that being said, it is not uncommon to see automakers outsourcing engines, transmissions, software; but you rarely see any major automaker outsourcing its body/structure design.
 
Well, there's not a lot you or I can do about the design of the vehicle frame. I found this post more useful to keep in mind for the long term. For example, if this torsional rigidity thing is true, it opens up a window for one of Tesla's "rivals" to make a better car. So far, nothing comes close to Tesla's products for most of us, so we live with the boom-boom.

If you believe Sandy Munro, it will be a Chinese manufacturer who pulls it off. NIO was mentioned specifically in a reply. I am waiting.
 
Musk sure seems to corroborate the OP’s sentiments in this interview with Sandy Monroe. Specifically the section discussing structural battery packs around the 22 minute mark.

Musk Monroe interview

He essentially admits the cars have poor torsional rigidity due to the battery packs and that resolving this would alleviate the NVH issues.

Maybe once the Y ships with the 4680 cells in this new honeycomb packaging it’ll be better.
 
So far Tesla is the only company that has the money to afford and engineering know-how to do megacastings. They're going to save a lot of money on manufacturing which will result in a better vehicle. This entire thread sounds like FUD because the competition can't afford megacastings.


I'd agree with Sandy, but your interpretation is wrong/BS/FUD. Watch this video and you won't see the glass roof cracking ("this is the reason why we see so many glass roof crack complaints").


People have been racing the M3 on tough tracks that put a lot of load on the car and it's solid. MY F & R mega castings take that a step above M3. Castings are so rigid they're generally brittle.

Musk sure seems to corroborate the OP’s sentiments in this interview with Sandy Monroe. Specifically the section discussing structural battery packs around the 22 minute mark.

Musk Monroe interview

He essentially admits the cars have poor torsional rigidity due to the battery packs and that resolving this would alleviate the NVH issues.

Maybe once the Y ships with the 4680 cells in this new honeycomb packaging it’ll be better.
 
Yeah, that "so many glass roof crack complaints" line got me too. I find a handful when I search, but it's not clear that any of them are from torsional stress.

Secondly, while the torsional mode may have a fairly high peak in the structural analysis, the coupling to acoustics is small. You're talking about 30 Hz and below, so the wavelength is >> the cabin size, which means you have to look at the net compression of the air from all the motions. Twisting the cabin has zero first-order effect on the air volume, so the only coupling is higher-order. Seems to me you'd need really major torsional amplitude to get a strong acoustic signal.

I would venture that the details of how air gets in and out of the cabin is much more important for low frequency noise. You're basically sitting inside a Helmholtz resonator. The air mass in the vents, and the amount of damping there, is going to make a big difference in how much excitation you get in the car at those frequencies.
 
So far Tesla is the only company that has the money to afford and engineering know-how to do megacastings. They're going to save a lot of money on manufacturing which will result in a better vehicle. This entire thread sounds like FUD because the competition can't afford megacastings.


I'd agree with Sandy, but your interpretation is wrong/BS/FUD. Watch this video and you won't see the glass roof cracking ("this is the reason why we see so many glass roof crack complaints").


People have been racing the M3 on tough tracks that put a lot of load on the car and it's solid. MY F & R mega castings take that a step above M3. Castings are so rigid they're generally brittle.

It takes certain rounds (thousands? hundred thousand?) of the stress cycle with enough stress level to make the crack occurred in the glass ceiling. That is the reason why not all of us observe this symptom in our cars, or in one single specific diagonal test. And it is possible that Tesla has changed the bonding formula which has more elasticity between the car frame and glass, to not let the glass ceiling be responsible for load, in exchange for less torsional rigidity.

Global torsional rigidity is mainly for NVH, ride comfort and ease of drivability, to tackle challenges in public roads/highways which are generally lack of proper maintenance.

So counter to your intuition, global torsional rigidity is less important for race track usage, since there is almost no damaging pothole, frost crevices, expansion joints/bumps there (also, race track surfaces usually is "flatter" than public roads). With that being said, there are significantly fewer chances for a vehicle to experience real challenges in global torsional rigidity on race tracks, compared with driving on public roads. Instead, front/back bulkhead local rigidity is more important for racing because you want to maintain a stable geometry of your front/rear suspension, so you see racing teams typically add strut tower bars, or change rear subframes, which strengthens front/rear bulkhead local rigidity only. This is also the reason that some racing-oriented sports cars (for example, targa type structure) have less global torsional rigidity (because they do not need it), to name a few: Corvette (C7 gen only has 14,000 nm/deg), Viper (7,600 nm/deg) etc.
 
I adjusted my hatch, lowered my 19" tire pressure to 38, added some sound deadening material (1/4 panel cavity made the biggest difference) and the road noise/wind noise gaskets. My car is quiet and pleasant. By far, the most important thing is for the hatch adjustments to be just right.
 
  • Like
Reactions: gundarx
Your FUD is weak and isn't going to work... Only Tesla is building their vehicles with mega-castings. Any new EVs that don't have megacastings are outdated and will suffer on price/margin and performance.

People have been racing Tesla M3s everywhere, including tracks that beat-up on cars like Nurburgring and Laguna Seca where there's big elevation changes, you're smashing over berms, and diving through massive banked corners at races speeds. Tires, brakes, wheel bearings won't last, but all of the glass stays intact ("this is the reason why we see so many glass roof crack complaints"). :D



It takes certain rounds (thousands? hundred thousand?) of the stress cycle with enough stress level to make the crack occurred in the glass ceiling. That is the reason why not all of us observe this symptom in our cars, or in one single specific diagonal test. And it is possible that Tesla has changed the bonding formula which has more elasticity between the car frame and glass, to not let the glass ceiling be responsible for load, in exchange for less torsional rigidity.

Global torsional rigidity is mainly for NVH, ride comfort and ease of drivability, to tackle challenges in public roads/highways which are generally lack of proper maintenance.

So counter to your intuition, global torsional rigidity is less important for race track usage, since there is almost no damaging pothole, frost crevices, expansion joints/bumps there (also, race track surfaces usually is "flatter" than public roads). With that being said, there are significantly fewer chances for a vehicle to experience real challenges in global torsional rigidity on race tracks, compared with driving on public roads. Instead, front/back bulkhead local rigidity is more important for racing because you want to maintain a stable geometry of your front/rear suspension, so you see racing teams typically add strut tower bars, or change rear subframes, which strengthens front/rear bulkhead local rigidity only. This is also the reason that some racing-oriented sports cars (for example, targa type structure) have less global torsional rigidity (because they do not need it), to name a few: Corvette (C7 gen only has 14,000 nm/deg), Viper (7,600 nm/deg) etc.
 
  • Like
Reactions: AlexHung
I adjusted my hatch, lowered my 19" tire pressure to 38, added some sound deadening material (1/4 panel cavity made the biggest difference) and the road noise/wind noise gaskets. My car is quiet and pleasant. By far, the most important thing is for the hatch adjustments to be just right.
If you don’t mind, please provide more details on what you have done, especially on the deadening of 1/4 panel cavity. Thanks a bunch.
 
Your FUD is weak and isn't going to work... Only Tesla is building their vehicles with mega-castings. Any new EVs that don't have megacastings are outdated and will suffer on price/margin and performance.

People have been racing Tesla M3s everywhere, including tracks that beat-up on cars like Nurburgring and Laguna Seca where there's big elevation changes, you're smashing over berms, and diving through massive banked corners at races speeds. Tires, brakes, wheel bearings won't last, but all of the glass stays intact ("this is the reason why we see so many glass roof crack complaints"). :D



I reviewed your post and the videos you provided as examples, one common knowledge missing in your argument is, on race tracks, a vehicle approaches the berms from the tangent direction relative to the berm curve, this means the real speed the vehicle "hits" the berm (vector component which is perpendicular to the direction of travel) is in fact very slow.

In other words, if a vehicle hits the berm in a way that causes large enough torsional stress to the car frame, then it means the vehicle has already been in a state of excessive understeered (or oversteered) in this cornering, which means an almost sure crashing event.

With that being said, this is the reason why so many sports cars using the Targa type structure, and their designers do not even bother with fiddling with the torsional rigidity too much.

Also, cornering itself will not cause serious torsional stress because it is the whole side of the vehicle (2 outer wheels) taking the load.
 
If you don’t mind, please provide more details on what you have done, especially on the deadening of 1/4 panel cavity. Thanks a bunch.
See: Ear pain/Pressure help

Insulating the bucket and hatch may not be worth the effort.

One more thing of note, after fixing the hatch (car was unbearable with the hatch buffeting), the car seemed to get quieter over the first few thousand miles. I'm not sure if the suspension became a little more compliant as it broke in, the tires broke in, door and hatch gaskets settled in, or something else.

I noticed the low frequency resonating was louder the other day, so I lowered my tire pressure back to 38 and it's mostly gone again. (Cold tire pressure crept up to 42 because of the warmer weather.)
 
Last edited:
Your FUD is laughably weak and isn't going to work... Only Tesla is building their vehicles with mega-castings. Any new EVs or normal ICE vehicles that don't have megacastings are outdated and will suffer on price/margin and performance. All Teslas are top-notch high-performance vehicles and performance will only increase with megacastings reducing weight.

Everyone drives over speed bumps or curbs to get into drive ways. Front tires hit the speed bumps before the rears. Many times you don't hit perfectly square and one side hits before the other. Roads aren't flat and front/rear and left/right experience different loads all the time.

Those Teslas drivers were putting enormous loads (torsional or whatever) on their vehicles when they smash over curbs or dive into catch berms at high-speed (100+ mph). The glass roof stays intact ("this is the reason why we see so many glass roof crack complaints") so your entire premise is complete nonsense.

I reviewed your post and the videos you provided as examples, one common knowledge missing in your argument is, on race tracks, a vehicle approaches the berms from the tangent direction relative to the berm curve, this means the real speed the vehicle "hits" the berm (vector component which is perpendicular to the direction of travel) is in fact very slow.

In other words, if a vehicle hits the berm in a way that causes large enough torsional stress to the car frame, then it means the vehicle has already been in a state of excessive understeered (or oversteered) in this cornering, which means an almost sure crashing event.

With that being said, this is the reason why so many sports cars using the Targa type structure, and their designers do not even bother with fiddling with the torsional rigidity too much.

Also, cornering itself will not cause serious torsional stress because it is the whole side of the vehicle (2 outer wheels) taking the load.