Reader map
What this page proves
Vibration transmissibility from repeated A/B runs. Start here for the build, question, measurement, and closeout proof.
What we are making
Two PETG phone mounts: one rigid and one rubber-grommet isolated, both on the same seat-rail adapter plate.
Question answered
Do rubber grommets reduce phone acceleration on the same route, and at which frequencies?
What will be measured
Static sag, tap-test frequency, route acceleration, RMS shaking, and spectral ratio across A1, B1, A2, B2.
Proof that closes it
05_bracket_A_photo.jpg, 05_bracket_B_photo.jpg, 05_mount_drawing.pdf, 05_static_sag_table.csv, 05_tap_test_raw.csv, 05_route_log.csv, four raw run CSVs, 05_rms_table.csv, 05_spectral_ratio_chart.png, and 05_summary.pdf.
Rubber can isolate a phone from fast shaking, but it can also bounce at its own favorite frequency. This test compares the same phone on the same route with and without rubber.
Words you need
Project terms
| Term | Meaning on this page |
|---|---|
| Vibration | Repeated shaking. |
| IMU | Phone motion sensor package. |
| Accelerometer | Sensor measuring acceleration. |
| RMS | Average shaking level. |
| Spectrum | Vibration by frequency. |
| Natural frequency | Frequency where mount wants to bounce. |
| Transmissibility | Output/input motion ratio. |
| Grommet | Rubber fastener isolator. |
| Stiffness | Force divided by deflection. |
| Tap test | Hit used to find ring-down frequency. |
| High-pass filter | Removes slow tilt/gravity drift. |
Capstone framing
Objective, requirements, constraints
| Type | Specific requirement |
|---|---|
| Objective | Compare rigid and isolated mounts. |
| R1 | Two brackets from one controlled design. |
| R2 | One phone/orientation for every run. |
| R3 | Fixed sample rate >=100 Hz. |
| R4 | At least two A/B route pairs. |
| R5 | Tap-test natural frequency <=25 Hz. |
| R6 | RMS and spectral ratio published. |
| Safety | Adult/builder approves mount and clearances. |
Section 01 · Concept
Experiment design
Only one thing differs between the two brackets: the four grommets. Everything else — the mounting plate, the phone, the clamp tightness, the route, the day — stays the same. That way, any difference between the recordings can be traced to the rubber.
What it is
Single-variable experiment
Two PETG brackets print from one CAD model and bolt to a shared plate on a 40 × 40 mm four-bolt pattern. That plate slips under the car's existing seat-rail bolt, so no new holes go in the car.
Version A is rigid; version B sits on four rubber grommets — the one change under test. Each bracket gets at least two 10-minute drives, recorded by the phone.
Because only one thing changed, the grommets are the first explanation for any difference. Driving the route at least twice per bracket shows how much the runs vary on their own. If the rubber's effect is smaller than that natural variation, the report says so — that outcome is still useful.
Actual principle used
Vibration transmissibility from repeated A/B runs
Three ideas run through this build: soft mounts that filter shaking, a fair one-variable comparison, and a standard way to average vibration (RMS). Each one shows up as a deliverable in Section 05.
Isolation
4 grommets
Control
1 variable
Metric
RMS x · y · z
Evidence
6 items
Section 04 · Sequence
Build sequence and pass/fail checkpoints
The build runs in eight steps: measure, design both brackets, print, assemble, install, drive, and compute. Six pass/fail checkpoints keep the comparison fair; the project is done when all six are met and the evidence folder ships.
Sequence
Build sequence
Measure. Caliper the seat-rail attachment point and bolt clearances; weigh the phone-plus-clamp stack the grommets will carry.
Design A. CAD the rigid bracket and phone-cradle interface on the shared 40 × 40 mm four-bolt pattern fixed at the Measure step.
Design B. CAD the isolated bracket with four grommet seats — 8 mm panel holes in a 3 mm flange — plus clearance for the brass sleeves.
Print. Produce both brackets in PETG — 235–245 °C nozzle, 75–85 °C bed, 0.2 mm layers, 4 perimeters, 40% infill, bolt bosses oriented so layer lines run across the bolt axis — then heat-set the M4 inserts into 5.6 mm pilot bores with the soldering iron, pressed square until flush.
Mount. Fit the phone in one orientation, cradle screws set to 1.2 N·m into the inserts; attach the tether routed clear of the controls.
Install. Bolt bracket A to the rail plate — 4× M4×20 with fender washers and nylocks at 1.5 N·m by torque driver — and hand-check deflection: pass is no perceptible motion on the rigid mount.
Drive. Log run A over the 10-minute route; park, swap to bracket B with the same four bolts torqued to the same 1.5 N·m against the brass sleeves, tap-test the ring-down (25 Hz or below to pass — fit softer grommets and re-test otherwise), hand-check deflection (~1–2 mm springy travel, no bracket-to-plate contact), log run B; repeat the pair so each bracket has at least two runs, exporting raw CSV while parked.
Compute. Trim all logs to identical duration, remove the per-axis mean (or high-pass at 1 Hz) to strip gravity, compute per-axis RMS and Welch spectra, chart the isolated/rigid ratio against the predicted crossover, photograph both installs, and record one design change.
Pass/fail checkpoints — owner: Kohler
Pass/fail checkpoints
Risk, judging, and closeout
What can fail and how the result is judged
What could go wrong
| Risk | Control | Failure action |
|---|---|---|
| Unsafe location | Check pedals, steering, airbags, seat travel. | Move mount or stop. |
| Seat-rail misuse | Use service torque if touched. | Do not drive until signed off. |
| Grommets too stiff | Static sag and tap test. | Change grommets. |
| Orientation change | Mark phone and clamp. | Discard run. |
| Route variation | Run A1, B1, A2, B2 same day. | Repeat pair. |
| Sensor clipping | Check raw range. | Repeat. |
How the result will be judged
| Check | A-level result | Not acceptable |
|---|---|---|
| Fair control | Only grommets change. | Different route/phone/clamp. |
| Frequency check | Tap test <=25 Hz. | No frequency evidence. |
| Data depth | RMS plus spectrum ratio. | Subjective less shaky. |
| Repeatability | A1/B1/A2/B2 consistent. | One lucky run. |
Data package
Exact closeout filenames
- 05_bracket_A_photo.jpg
- 05_bracket_B_photo.jpg
- 05_mount_drawing.pdf
- 05_static_sag_table.csv
- 05_tap_test_raw.csv
- 05_route_log.csv
- 05_run_A1_raw.csv
- 05_run_B1_raw.csv
- 05_run_A2_raw.csv
- 05_run_B2_raw.csv
- 05_rms_table.csv
- 05_spectral_ratio_chart.png
- 05_summary.pdf
Senior capstone readiness
Current status: build plan; prototype and data pending. A-level requires stiffness characterization, transmissibility prediction, repeated runs, uncertainty/repeatability, and CAD-to-test comparison.
Definition of done
- Both brackets and drawing are published.
- Static sag and tap test are recorded.
- A1/B1/A2/B2 raw files are preserved.
- RMS table and spectral ratio chart state the result.
Section 02 · Method
Isolation principle
Vibration reaches the phone through the bracket; the grommets add a compliant layer in that path. Standard transmissibility sets the cutoff: isolation helps only above about 1.4× the mount's natural frequency, and amplifies near it. The rest of the method holds every run to identical conditions.
Bracket pair on the shared rail · live model
One seat-rail attachment, two bracket versions: rigid bolt-up at left, four-grommet isolation at right. Phone and clamp stay identical across every run.
Design rules
Target natural frequency: 15–25 Hz
suspended mass ≈ 0.35–0.45 kg (phone ~200 g + clamp ~100 g + isolated bracket top ~80 g) — weigh the stack, then pick stiffness so fn ≈ 15–25 Hz
that target needs k_total ≈ 3.4–9.4 kN/m (0.9–2.4 kN/m per grommet); hardware-store grommets in compression run 10–100× stiffer → spec soft silicone, 40–50 Shore A, or load the grommets in shear, and confirm fn by static sag (0.4–1.1 mm expected) or tap test
crossover fn·√2 ≈ 21–35 Hz → content above that band should attenuate, while wheel hop (≈10–15 Hz) and idle firing (≈20–30 Hz) sit near fn and will be amplified — the A/B spectra are expected to show both effects, and that is the test
Each run boils down to three numbers: the average shaking left-right, front-back, and up-down (RMS). A frequency chart then shows where the rubber helps, compared to the predicted 21–35 Hz crossover point. One detail protects the whole comparison — brass sleeves inside the grommet holes keep the bolts centered, because a loose bolt would rattle and corrupt the data. Phone sensors make this a trend-grade result: good enough to say which mount wins and by how much.
Controls, fixed before the first run
Shared mounting plate
Both brackets bolt to the same plate, so the path into the car is shared hardware. Every run uses one phone, one orientation, one clamp tightness, one 10-minute loop, one driver, and one day, with tire pressures checked before the first run.
Logging starts and stops only while parked, and the phone's tether stays clear of the pedals and wheel. A free app (SensorLogger or Phyphox) records at one fixed rate — 100 Hz minimum — and every file is saved as raw CSV before any math is done. Before driving, each mount gets a push-by-hand check.
Attachment
1 rail plate
Phone
same unit
Torque
1.2–1.5 N·m, driver-set
Logging
raw CSV, one rate
Zoo / SolidWorks — build this model
Bracket models and frequency study
Both brackets exist as CAD before anything prints, so SolidWorks is already in this build's path. Model them as two configurations of one part: rigid, and grommet-mounted. A frequency study on the grommet version predicts the mount's own bounce rate — the number the 21–35 Hz crossover on this sheet is computed from.
Produce the dimensioned interface drawing named in the evidence folder: the 40 × 40 mm bolt pattern, the grommet bores, and the clamp face. STL files for the printer export from the same models, and the macro saves the drawing as this page's PNG. A working model of this build is embedded just below, and its STL download opens in SolidWorks, Onshape, or FreeCAD.
Runs in: Zoo (free — its Zookeeper agent builds from this prompt, edits by conversation, and answers design questions) · SolidWorks LEO · or by hand from the list above.
The Zoo file goes into the zoo-design-studio-projects folder shown at the top of Zoo’s Projects screen; the build then appears in the Projects list, dimensioned from this sheet. Paste the macro into SolidWorks (Tools → Macro → New) to export every drawing as a web image.
Model
2 configurations
Check
frequency study
Drawing
interface
Export
STL + PNG
If the frequency study is out of reach
Section 03 · Parts
Parts and cost
Hardware totals roughly 160 g of 3D-printing plastic (PETG), a handful of small M4 bolts and nuts with brass sleeves, and four rubber grommets. A drilled aluminum plate rides under the vehicle's existing seat-rail bolt. Printer, phone, and vehicle are already owned, which holds the cash cost to $35–75.
Bill of materials
Parts and materials
| Item | Qty |
|---|---|
| Phone (one unit for every run) | 1 |
| Phone clamp or cradle | 1 |
| PETG bracket bodies, ≈80 g filament each | 2 |
| Seat-rail plate, 3 mm aluminum, drilled to suit — mounts under the vehicle's existing seat-rail bolt (OEM bolt reused, retorqued to the service-manual spec, typically 35–50 N·m) | 1 |
| M4×20 bolts (bracket-to-plate) | 4 |
| M4 nylock nuts (bracket-to-plate through-bolts) | 4 |
| M4 fender washers, 12 mm OD (spread load on the grommet faces) | 8 |
| Rubber grommets, panel-groove type, 6 mm ID × 10 mm OD, groove sized for a 3 mm flange in an 8 mm hole; soft silicone, 40–50 Shore A | 4 |
| Brass shoulder sleeves, 4 mm ID × 6 mm OD × 5 mm — center the bolts in the grommets and carry the preload (≈$5) | 4 |
| M4 threaded inserts, heat-set (phone-cradle bosses in the print) | 4 |
| Cable tether | 1 |
| Parts total | $35–75 |
Tools
Tool list
CAD software and a 3D printer make the brackets. Calipers size the rail attachment and the bolt holes; an M4 hex key does the assembly. A drill opens the plate holes, a soldering iron with an insert tip sets the brass inserts, and a small $20–40 torque driver makes the clamp tightness repeatable instead of guessed. The phone app records every drive.
Time
20–30 h
Cost
$35–75
Filament
≈80 g per bracket
App
SensorLogger or Phyphox
Section 05 · Scope
Scope and next steps
A phone sensor is enough to show which mount wins and by how much, with raw logs kept for reanalysis. A calibrated accelerometer on the same brackets is the step to lab-grade numbers. The transmissibility physics is the same at full scale.
Stated scope
Instrumentation grade
A phone's sensor is not a lab instrument, and the report says so next to the charts. What it can do is show a repeatable difference between the two mounts. Bolting a calibrated accelerometer to the same two brackets is the follow-on: identical fixtures, lab-grade sensor, publishable numbers.
Evidence package
Six deliverables
One folder ships it all: both printed brackets installed, the raw sensor logs from every run, the vibration chart, and the install photos. The CAD files go in too, with one dimensioned drawing and a note recording one design change made along the way.
Where it leads
Connection to full-scale mounts
The rubber-mount math used here is the same math that decides how the jet-engine stand (build 01) bolts to its bench, and how turbine test stands are mounted after that. One rule — soft mounts help only above 1.4 times the mount's bounce rate — scales from a 0.4 kg phone bracket to test-stand hardware.