Why Engineers Use $1000 Hammers To Hit Prototype Cars So They Sound Good And Don’t Explode

Nvh Ts
ADVERTISEMENT

As a noise, vibration, harshness (NVH) engineer, I get to play with some neat tools. From lab-grade microphones and accelerometers, anechoic chambers, acoustic cameras, to high tech data acquisition systems. This equipment is important, as cars and trucks are complex machines in a market that is more competitive than ever. Despite thousands of explosions occurring in the engine every minute, whirring harmonics from gears and bearings, the road battering the suspension, and wind rushing across and through the body, your vehicle’s interior must be a pleasant place to be. Manufacturers aren’t just looking for quietness; the sound quality you experience from when you slam the door to when you rev the engine must match the character of the brand and vehicle you are driving.

The tool-of-the-trade that surprises most people is the modal hammer. This isn’t something you can buy at Home Depot, it’s a precision-tuned instrument that costs thousands of dollars. Every vehicle, from the complete system down to the component level has been hammer tested throughout its development. The data gathered while tapping away with these hammers not only ensures that your vehicle sounds and performs as it should, it also confirms that it won’t fail in a catastrophic explosion. Here’s how.

[Editor’s Note: Everyone please welcome Steve Balistreri, a fellow enginerd and Detroiter. He’s an expert on NVH, and since I’m fascinated by the topic, you can expect to see lots more from him soon! -DT]. 

Screenshot 2023 09 08 At 9.22.22 Am Copy
Bruel & Kjaer

A huge aspect of nailing the NVH performance of a vehicle is managing resonances. An example you are familiar with is how your vehicle’s mufflers are tuned to cancel out specific frequencies your engine makes. There’s a reason deleting your muffler often creates a loud drone at common cruising speeds; my 2001 S8 with a rear muffler delete is a prime example that gets really loud right at 45 mph. The OEM mufflers were tuned deliberately for this.

Resonance Can Make Cars Ridiculously Loud And Uncomfortable

What do I mean by “resonance”? Well, everything has frequencies that it naturally wants to resonate at. Body panels, subframes, engine blocks, driveshafts, the air cavity in your vehicle’s cabin (the Supra wind buffeting issue is a prime example of this). In the case of musical instruments, they are tuned to resonate at specific frequencies or notes that are pleasing to the ear. The components of a vehicle are specifically tuned to avoid frequencies they are likely to be excited by, such as engine firing, gear harmonics, or forced inputs from your vehicle’s suspension. The reason is best shown in the animation below, plotting vibration amplitude vs frequency.

You can see that below the *resonant frequency* the response is pretty flat, meaning the part will vibrate about the same amount as what is put into it. However, as you approach the resonance, amplitude increases exponentially to the point where your part is vibrating at 10x the amplitude as the vibration you are putting in. As you can imagine, this is a bad thing.

University of Utah

A classic example is the opera singer breaking a wine glass with her voice. The physical force from the sound of her voice is miniscule, and the glass is unaffected (unless of course the glass is some sort of opera connoisseur). Once that special frequency is hit, the once-rigid glass starts undulating in impossible ways until it finally explodes. The video below from the excellent Slow Mo Guys channel is a great visual. Trying to deform a wine glass like that with your hands would be impossible, but a speaker at the right frequency can turn that rigid glass into jello.

Screenshot 2023 09 08 At 9.23.56 Am

In a car this can cause a few things to happen. Most often a specific gear will suddenly get loud at a certain engine RPM range as a mount or subframe resonates at that gear’s frequency (the frequency of the powertrain depends on the engine RPM, which is set by the gear when traveling at a steady highway speed), or your car can sound very “boomy” as different road surfaces or engine frequencies excite resonances in your vehicle body.

Resonance Can Literally Cause Things To Explode


Other times this concept of resonance can cause catastrophic failure. A common example is driveshafts. Some vehicles are speed-limited so that a once-per-rotation vibration caused by imbalance doesn’t intersect with the driveshaft’s first resonant frequency. There were several instances of V6 Mustang owners deleting their speed limiter, then getting the rude surprise of their driveshaft exploding as they flew down the drag strip at over 130 mph. There are videos on YouTube of boneheads doing this on public roads. As you can hear, it’s a catastrophic failure (and you can see the aftermath here).

Screenshot 2023 09 08 At 9.27.19 Am

OK But How Do You Prevent These Resonance Problems?

The company I work for, HBK, is based in Denmark and produced training materials in the 80s with delightful hand drawn illustrations. Here’s one showing an example of how resonances can amplify frequencies you don’t want. In the bottom left graph (commonly called a waterfall graph) the diagonal “waves” are engine orders (firing frequencies based on the number and layout of the cylinders that increase with RPM). That engine order crosses a 30 Hz resonance at 100 kph, where the acceleration (Z axis) increases dramatically. The top graph shows that engine order sliced out and plotted as acceleration (vibration) vs speed. As you can see the resonances dramatically increase the level of vibration.

Screenshot 2023 09 08 At 10.16.30 Am
Bruel & Kjaer

But how do you calculate resonant frequency? If you went to engineering school prepare to have some flashbacks. Below is a simple mass / spring diagram which represents the simplest, ideal structure. The resonant frequency equation is below and is based on the mass m and spring constant k (or stiffness in a non-spring), and these are what you need to change if your resonant frequency is in the wrong place.

Think of stretching a guitar string to change its pitch. This is a very simplified example, like comparing the driving in Police Quest with Forza. In the real world, parts have internal damping inherent to their structure, the shapes themselves are complex, and the equations get more complicated. Adding damping flattens the peak and lowers the frequency slightly. You can imagine this like a tuning fork which has very low damping; it will ring forever at a very sharp tone if you tap it with a metal rod. While if you tap your desk with a metal rod, it’s a much broader set of tones that go away almost immediately. This is the difference damping makes.

Screenshot 2023 09 08 At 10.18.58 Am
Georgia State University

Let’s Talk Hammers

All this is modeled out in CAE/FEA (Computer Aided Engineering/Finite Element Analysis) software early in the vehicle program. Engineers do what is called modal testing with these hammers on prototype parts to help correlate these models and confirm these frequencies. But why do we need a hammer that costs as much as a used car? Can’t you use a regular hammer, your knuckle, or a petrified French fry you found between the seats, and just record the sound with a microphone? Technically you can, and sometimes for quick and dirty troubleshooting we may just tap around to find a part that rings around the problem frequency, but that only gets you half of the equation.

To get the frequency response or sensitivity, you need to know your input and your output. You can tap a driveshaft with your knuckle all day, but how much force are you putting in to get it to ring that amount, and what frequency range are you exciting? These hammers have force transducers built into the business-end that records the force you are inputting into the system to get that input to output ratio. They also come with interchangeable tips with different hardness to excite different frequency ranges. For example, tap your coffee cup with the fleshy part of your finger and again with the tip of your nail. Sounds different right? Soft rubber tips will set off any low frequency modes, while harder nylon and even steel tips put high frequency energy into the system.

Having the force and response gets you a Frequency Response Function like you see below, displaying how much vibration you get per Newton of input force vs frequency. The peaks are the resonant frequencies. You usually test multiple points on the part which is why there are several different traces on the graph.

42452 2019 806 Fig6 Html
Springer Nature

The hammers come in different sizes depending on the part being tested. There are tiny baby hammers smaller than a Bic pen for testing small parts like hard drives, and large sledge hammers for giant structures like bridges or the NASA launch pad I got to do structural testing on. The 8206 hammer is our workhorse (note, images below are not to scale).

Hammers Copy
Bruel & Kjaer

Here is a photo of a modal test being done on a car mirror. You can see the accelerometer is attached to the mirror glass. They are probably confirming that the parts in the mirror won’t be excited by vehicle vibrations, which would cause the mirror to shake excessively, blurring the reflection and causing a safety issue.

Impacttesting Rearviewmirror
M+P International

The hammers are also modally tuned, so the hammer itself doesn’t have resonant frequencies in the range you are testing. You want the part to be as isolated from outside interference as possible, so if the hammer was resonating along with the part, you’d have bad data. This is also why we isolate the parts as much as we can, suspending them with bungee cords if they are small enough.

Large parts like vehicle bodies get suspended from the ceiling with rubber straps or are mounted on airbags to isolate them from the ground. It is an odd sight walking into an anechoic chamber and seeing a full pickup body, instrumented with hundreds of accelerometers and strain gauges, hanging from the ceiling with some poor engineer underneath hitting it with a hammer. It’s a tough job, but somebody has to do it. Below is another drawing from our company’s training manual showing a vehicle body being isolated from its environment. In this photo they are using a shaker instead of a hammer but it’s the same idea.

Screenshot 2023 09 08 At 10.39.42 Am
Bruel & Kjaer

So the engineer will map out several points of interest on the part in question to glue accelerometers to, and then do a modal survey (basically tap the part in a few different spots to see what works best). They will then tap that special spot with their hammer, which will trigger the data acquisition software to record the hammer force and the resulting vibration from the accelerometers. Typically, it’s five hits per location to average out any inconsistencies. The hits must be quick and clean — no double rebound hits or any scraping movements, because it messes up getting that clean initial impact. This can be tough when your arm is snaked in an engine bay, hitting some transmission mount you can’t even see. You don’t need a whole lot of force to do this — light taps will do. There is a story of a new guy who was testing a car’s roof and was hitting it hard enough to dent the roof. Not only does that give you bad data, it screws up the hammer. He learned his lesson pretty quick.

An example test is shown in the illustration below. The accelerometers record the resulting vibration from your hammer hit, the data goes into our giant mainframe, tape reels do whatever they do, and it spits out the results. We print the graph out with our extremely loud dot matrix printer, give it to the head engineer and he gets mad.

Screenshot 2023 09 08 At 10.41.01 Am
Bruel & Kjaer

For a full modal test, the measurement point coordinates are imported into our software to create a wireframe model. The test data is then imported and processed, and we get the resonant frequencies as well as the mode shapes. The mode shape is like the shape the wine glass was deforming into during the above video, each frequency has its own shape, and they get more complex as you increase the frequency. An example of a vehicle mode shape is below (the colorful one). This is CAE data; our experimental data has significantly lower resolution as it’s just a wireframe of a few dozen/hundred points, more like the image below that, but you get the idea.

Screenshot 2023 09 08 At 10.42.44 Am
Bruel & Kjaer

 

EMFAD

A Look At Data On A Pickup Truck Body

Below is a screenshot of our analysis software doing some curve fitting and extracting mode shapes from a test on a truck body. As you can see, there are several modes between 12 and 36 Hz. The mode shape at 17.6 Hz is shown in the geometry window and appears to be the rear of the truck bed wagging back and forth like a dog’s tail.

Advanced Modal Analysis 1
Bruel & Kjaer

The mode shape in the CAE animation above most likely happens at a very low <40 Hz frequency and would be heard and felt as a booming noise over certain road surfaces or bumps. It may be hard to believe that your heavy steel vehicle body is wobbling like a giant Jello jiggler but that is actually happening. The part needs to vibrate in some way to create the sound waves and vibration that eventually reaches your ears and your butt. The displacement in these animations are greatly exaggerated so you can visualize what the shape is, as it is typically invisible to the human eye, but not always.

We tested a prototype industrial engine which had a fuel line that would resonate around 450 Hz, vibrating so much it would look blurry at high RPMs. We glued an accelerometer to it and measured a whopping 122 g’s of vibration! As a reference the engine block rarely exceeds 2 g’s. At 450 Hz that fuel line is moving almost 300 millimeters per second which is pretty incredible. This would quickly fatigue the line, causing fuel leaks and fires in their engine dynos. We fabricated some brackets that increased the stiffness, shifting the resonance up and away from where it would be excited by the engine, which dramatically reduced the vibration and fixed the problem.

Automotive engineers have gotten a lot better at NVH over the past decade or so. Much of it is due to advances in CAE, and automakers making NVH considerations much earlier in the design process. There was a running joke that NVH stood for “not very helpful.” Early prototype vehicles are rattly, loud pieces of junk, so NVH concerns didn’t make themselves known until engineers got their hands on the later, more refined prototypes when most tooling was finalized. So solutions to NVH problems were either to lessen the source excitation ie. make the combustion process less explodey or the gears less geary (not really feasible); redesign part in question (extremely expensive and time consuming after the part/tooling has been finalized); or treat the path to the driver which involves adding mass, stiffness, or damping treatments (like tuned mass dampers) which all add weight and cost. Not what a manager wants to hear when they are looking for an easy fix. Thankfully these surprises late in vehicle development are becoming more rare as good NVH has become a higher priority.

Screenshot 2023 09 08 At 10.56.01 Am
Bruel & Kjaer

Fascinating Ways To Masking Problems Late In The Vehicle Development Timeline

In the modern days auto companies have gotten creative for masking these problems when they can’t be bothered to fix them. One example is ANC or active noise control, which uses cabin microphones and your stereo speakers to cancel out problem frequencies. These work a lot like your noise canceling headphones but are only effective for low frequency boom sounds in your car. It works quite well and is much cheaper than significantly changing the vehicle body structure.

Another more humorous solution is to edit out specific RPMs, literally making them disappear. A common boom detection test is to do a slow (45 second) engine runup from idle to 1500 RPM while the vehicle is parked. This will cause the engine frequencies to excite any problem areas in the idle range. One particular car would not let me get a smooth ramp, the tach needle avoided 1100 RPM like a baby swatting away a spoonful of steamed spinach. I asked an engineer and it turned out the engine vibrations set off a resonance somewhere in the vehicle at this RPM, so they programmed it to simply skip past this engine speed so the resulting boom is barely noticeable. Once again adding a couple lines of code is much cheaper than changing the tooling for a major structural component, so they went for it.

Our future with EV’s presents unique challenges and opportunities. For one, you don’t have an engine, transmission, and driveshaft flooding the vehicle with vibrations and noise. The problem with that is engine noise helps mask a lot of wind, road, and accessory noise so any issues there become more prominent. Also the heavy engine block acts like an isolator for things like your A/C compressor and water pump, which now have to be mounted on an isolated frame of some sort. Electric motors and inverters create less NVH comparatively, but it is at much higher frequencies than an engine which creates its own host of problems. Some automakers are experimenting with different materials for things like motor mounts that are less sensitive to high frequency vibrations.

There is also the problem of a lack of vehicle character, which relied heavily on the noise and vibration of the powertrain. Besides certain restrictions from federal pedestrian safety noise requirements, automakers have a blank slate to create an artificial noise profile that imparts a character to the vehicle. This takes a lot more work, getting into psychoacoustics and sound quality and perception for different customer groups, in my humble opinion we have a way to go with that. We will save that for a future article.

So next time you go for a drive, enjoy the solid thunk of your car door closing, the lack of loud engine booming, intrusive gear noise, buzzing interior bits, shaky mirrors, and the absence of exploding parts. Thank the humble NVH engineer, who spent days holed away in an anechoic chamber covered in glue, tirelessly tapping away with their expensive hammer, tuning your vehicle like a musical instrument to be as safe and enjoyable as possible.

About the Author

View All My Posts

84 thoughts on “Why Engineers Use $1000 Hammers To Hit Prototype Cars So They Sound Good And Don’t Explode

  1. There’s this steel pedestrian bridge with wooden platform over a creek in my old neighborhood that would be set off by my 50-lb dog trotting over it, but not by me walking or running or any number of people on it at once. autumn-view-of-a-bridge-across-campbell-creek-anchorage-southcentral-FG6R4X.jpg (1300×956) (alamy.com) How you would go about testing that with the modal sledge hammer? Single taps in different places or do you have to try tapping at different frequencies?

    1. That’s a reason why when soldiers need to cross a bridge they stop marching in lockstep because the regular synchronized impacts from their feet can set resonances off.
      For a bridge we’d use a sledgehammer and test different points since it’s such a large structure.

        1. A great engine design searies of articles could be written from an engineers perspective telling the differences between different types of engines: hit or miss, 4 cycle diesel vs. 2 cycle diesel, diesel pre-chamber vs direct injection, common rail vs. PD, gas 2 vs. 4 cycle, Miller cycle, push rod vs. OHC, camless, rotary, radial, etc…

          Likewise seeing as ev is likely to take over passenger cars, an article or series of articles on different ev motor designs, radial flux vs. Axial flux, rare earth magnet vs. Electro magnet, etc…

  2. Is temperature taken into account? My old Focus would developed a buzz somewhere around the instrument cluster that happened very specifically (1) in fourth gear (2) around 2000 RPM (3) at or below freezing. ~2000 RPM and fourth gear happened to the speed limit on basically every road in my daily commute, so driving in winter was even more annoying than it normally is. Luckily, it could usually be silenced with a carefully calibrated smack to the top of the dashboard.

    I miss that car. The buzzing, not so much.

    1. We test cars down to -40 (I’m using scientificish Centigrade degreees, but it’s the same number in vague-o-unit Fahrenheit degrees too).

      We have a bigger tolerance for weird noises at low temperatures because it’s a relatively low occurrence use case plus, and I’m sure this is unrelated, the NVH guys get really cold.

    2. They can model temperature changes in CAE, so if we can confirm the models at room temp they should be accurate at other temps. As captain muppet says it’s not large use case, although I’ve done several idle tests where we’d start the car and sit very still (so the mic doesn’t pick up our movements) in a -20 degree chamber and wait until the car warms up so see how the car sounds / behaves. It’s not a fun test lol.

  3. When we found that some injector baffless of Titan IV rocket motors had cracked welds, a test was needed to find which baffles were defective. I recommended the “twang test”. Which was to be executed via a hammer. If the ballfle responded with “thunk”, it was defective. If it went ‘twang” it was good. The vibration was blocked by the cracked weld, so it thunked. If it twanged the baffle rang with a complete weld. The AF decided to use ultrasound inspection instead. Yeah, more accurate, but not as fun.

    1. I do a similar test with mugs and bowls in the kitchen. One that has a hidden crack will have a dull sound that doesn’t go DING or sustain as much as an intact one.

      1. That’s also the way to test grinding wheels. Not so much the fiber-reinforced angle grinder wheels, but you ALWAYS ring-test any non-reinforced wheels before installing. Having even a small 7-inch diameter grinding wheel come apart at 3400 RPM is an experience I don’t want to have.

    2. I recall a Samcrac YouTube video where he replaced the timing belt & apparently the tone of the belt when plucked was/is supposedly the Ferrari certified way to determine the correct tension on the belt

  4. This was a fascinating article!

    However, I cannot imagine spending any length of time in an anechoic chamber. Those things are freaky-deaky!

    I went in one on a job interview, and the person I was interviewing with told how if you stayed in one long enough, you would be able to hear your heartbeat and all kinds of bodily functions.

  5. So this was how engineers determined the exact amount of body adhesive to use in place of traditional sound deadening on my factory lightweighted car.

  6. Personally, I feel most of this IS unnecessary. Sure for a fuel line obviously it was so bad a human would just see the motion.

    Arguing the usefulness of the idea masks my belief that its penny-wise and pound foolish.

    There are tons of better things and low hanging fruit to make customers happy than hanging my shifter from bungee cords and smacking it with a hammer. The whole point is that I *like* the vibration of the engine I can feel from it.

    This is like baking a really horrible cake no one wants and then spending an inordinate amount of time proving that some random food coloring combination will make it more wanted than another color with too much data no one needed to collect.

    Is food color of cake important? Sure argue all day. Is is easier to just bake a cake we want and skip the whole process, spending the money on better ingredients and a chef who knows what people want is MUCH better use of the resources.

    I literally try to find the most direct raw NVH laden linkage to the road and am much happier for it.

    First build a car with specs we actually want first. I as a consumer dont want this going into a basic car that now costs $40,000. Great article but seesh are these tests a waste beyond basic safety.

    1. I had a 4Runner with no carpet, mud terrain tires, and the balance shafts deleted (3RZ engine). 5 minutes in the passenger seat and you’d prefer to be walking.

    2. You sound like the kind of consumer who would be happy driving the Flintstone-Mobile™. As such, your opinion is in the extreme minority and thus is not helpful.

    3. I drive prototype cars a lot, and some of the NVH issues are horrific. You say you like the engine and transmission vibrations, but you’ve only ever experienced the ones that have been tuned to be enjoyable, or at worst not actively unpleasant.

      Plus a lot of this work isn’t for you directly, we had a load of misfires that we really struggled to find a reason for, and it turned out that little bracket on the inlet manifold was causing 50g vibrations of the throttle position sensor. Nothing the driver could ever feel.

      I’ve had a few disagreements with NVH about whether certain noises are desirable or not, but resonance is never good.

    4. > The whole point is that I *like* the vibration of the engine I can feel from it.

      I doubt you’d enjoy a car with no NVH engineering at all.

    5. There is a big difference between “good” NVH and “bad” NVH. Low frequency booming can get extremely fatiguing, there are many components that make noise that don’t add to the experience like from your AC system, whistles and air rush noise through the heating ducts, blade pass frequencies from the blower fan for example. Mirrors whistling from wind noise, etc etc I agree that modern cars can use more “good” NVH but no one wants the bad.

  7. Well Albert Einstein, Nikola Tesla, and String Theory say ALL is vibration. Good Grief Man, are you endeavoring to have the universe blink out of existence ?!

  8. Owned a 95 Infiniti G20. The engineers did an awful job, every acceleration between, what I recall, as 20-30 MPH, the front of the windshield header and rearview mirror would vibrate violently. It was fantastically annoying. Fascinating article, thank you.

  9. Chrysler Pacifica NVH engineers nailed it on the current 2017+ version. Well, I can’t speak for the earlier models by the 2022 has newer seals and acoustic glass in some areas. Combined with active noise cancellation that thing is amazingly quiet even with all seats folded. Keeping such a cavernous space so muted is quite impressive.

    1. I did some testing on those! It was an early hybrid prototype and didn’t like being on the 4 wheel dyno. ANC can do wonders for things like that, expect an article on that soon.

  10. Wow that was a great article. I’ve messed around with NVH in an aftermarket way doing my stereo upgrades, but you guys obviously take it to a whole new level.

    Frequencies rule the universe!

  11. I ran an antique car restoration shop for 25 years, and one of my specialties was the 1961-67 Lincoln Continental 4-door convertibles. The first time I removed the big chrome bumpers on my own 1966 Lincoln convertible, I found large cast iron weights hanging at the ends of heavy stamped steel brackets. The brackets were very thick & looked like typical leaf springs. They were mounted between the body frame and bumper brackets, with the weights suspended at the other end of the brackets. Since the sedans didn’t have these weights, I assumed it had something to do with the ride quality of the convertibles, so I left them on all the cars we restored.

    One of my customers brought his Lincoln convertible in to the shop, complaining about some of the doors opening to the first latch position [but not fully opening] when the car went over large bumps. When I put the car up on the lift, I noticed all 4 weights behind the bumpers were gone. He said that he removed them to cut down on weight because the car drank gasoline like crazy. I asked him to bring the weights and brackets in, and we re-installed all 4 sets.

    The first time we hit some potholes, the car’s unit-body wiggled just as they all do, but the door latches never popped open. That was my first experience with NVH.

    1. Great story! That is a common problem with convertibles. I remember seeing large loosely sprung weights while crawling around under a mustang convertible. Really low frequency stuff is hard to manage.

  12. The vehicle mode shape (the colorful one made from CAE data) is a 1995 Dodge/Plymouth/Chrysler neon, isn’t it? I know that body shell when I see it. I work as a technical training instructor for an automaker. One of the classes that I teach covers NVH diagnosis and repair. This article is Good Stuff! I did not know about the modal hammer.

  13. Now I want one of these hammers for scientific bonking.

    Am I the only one that was waiting for Tesla’s machanical oscillator to make an appearance in the examples, then was a little sad?

  14. Great article, thanks a lot. Made me recall what we were doing in the late 1970s and early 1980s to dress NVH issues on American production cars at the dealer warranty level. The process of remedy also required the use of hammers, sometimes powered by compressed air.

    The VW Rabbit, Dodge/Plymouth Omni twins, Ford Granada/Fairmont, GM X body all had major NVH issues at the time. As did countless other makes and models.

    We learned to isolate NVH noises on test drives, with the use of a length of rubber vacuum hose to search for the source as someone else drove. Then out came the hammers once the point of noise/vibration was found. One mechanic I knew would use a lift rack and wood blocks to dent the floors and fire walls until NVH was less offensive.

    A lot of NVH was generated within the dash area then as well. A long ratchet extension and hammer was the go to tool for this. The goal was to bend the steel behind the dash enough to quiet the noise.

    Thanks for the weird memories.

    1. That’s really funny. Putting a crease in a large panel like that will stiffen it up and break up the surface so it doesn’t act as a large speaker. It would have been tough as an NVH engineer back then. Computing power has come such a long way, also the amount of data we can take would have been unimaginable back then. Some of our recordings are several gigabytes, while back then they were using tape reels to record sound. Thanks for the story!

  15. This brings back some memories. My senior design project was for John Deere they were expereince excess vibration on the entry platform to one of their machienes. After looking at all sorts of things and doing all kinds of computer modeling with no clear results we finally stuck an accelerometer on the platform and wacked it with a hammer, the resonant frequency for the platform was identical to the idle frequency from the motor.

    The funny part was when we presented our finidngs and recommened an additional stiffner to reduce the vibration, they didn’t like the idea of added cost, so we came up with a way to make it weaker, lowering the frequency below the idle speed.

    1. We did a full vehicle modal test for another heavy duty off highway vehicle manufacturer. They had a very rudimentary “sound barn” that was right next to the main road. We were getting lots of random low frequency noise in the cabin. Turns out the large plexiglass bubble over the cab had a resonance right at a semi trucks idle frequency. When one was stopped at the light next to the barn the idle sound was actually much louder inside the cab than outside. Took us a while to figure that one out!

  16. This is a great article, thanks. Like everyone else, I came here for Jason and David. I find that I’m visiting every day, though, for this kind of stuff from Huibert, Adrian, etc.

  17. So next time you go for a drive, enjoy the solid thunk of your car door closing, the lack of loud engine booming, intrusive gear noise, buzzing interior bits, shaky mirrors, and the absence of exploding parts. Thank the humble NVH engineer

    To the contrary, the (controlled) explosions are the part I like best about driving.

    Even if the NVH engineers for some my cars made choices (or were directed to make choices) that I wouldn’t have preferred.

    Fascinating writeup regardless.

    1. Good news! In the same way you don’t need to know how to play the trumpet to annoy someone with it, you can just start removing things and eventually you’ll undo those pesky NVH engineers’ work!

Leave a Reply