REV O / MEASUREMENT DESIGN
From each specimen
to a complete test record.
The simulator demonstrates acquisition, calculation and trace retention. All generated forces and material responses are ARTIFICIAL. It has no connection to an Instron controller, Bluehill installation or physical robot.
Bi-directional three-point bending
There are three upper and three lower Ø10 mm rollers. The two outer pairs stay fixed, 75 mm apart. Instron drives the central upper/lower pair downward and upward. The board sits between each pair. The whole bending station remains raised by 150 mm; the head is already at the loading position before robot insertion.
| Item | Nominal review geometry | Effect on acquisition |
|---|---|---|
| Specimen model | 90 × 14 × 3.0544 mm | Width and thickness are assumptions from the existing model. |
| Central pair | 1 mm top and bottom pickup gap | Exclude pickup travel from loaded deflection. |
| Outer pairs | 7.800 mm opening | Configured rectangular-envelope take-up is 7.800 − 3.0544 = 4.7456 mm. |
| Illustrative method | Down → neutral → up → neutral | ±0.6 mm loaded deflection, +1.6 / −6.3456 mm head travel. |
| Lower center cassette | 50 mm rearward retract, positive mechanical lock | Both EXTENDED and LOCKED proofs required for bending. Empty/unloaded RETRACTED proof required for mode transfer. |
The upward phase first takes up the center gap and lifts the board toward the upper side rollers. The model generates no bending force during that free motion. The load train must be qualified for both directions, including the central return-pin carrier.
L = 75 mm; b = 14 mm; h = 3.0544 mm
F = ±10.8 N → σnom = ±9.3024 MPa
This is the elementary rectangular-beam outer-fiber relation. The sign follows the selected reference fiber and loading direction; opposite faces carry opposite stress. It is a nominal section estimate, not the local stress in chips, solder joints, multilayer PCB features or an SSD assembly. An approved section/FEA model or strain measurement is needed for those claims. Crosshead displacement includes load-train compliance and clearances; it is not automatically specimen deflection. CAD world +Y is upward, opposite to the downward-positive head-displacement channel.
Displayed channels: force–head displacement, nominal stress–head displacement, nominal stress–time, head displacement–time and force–time. The CSV separately retains corrected load-point deflection and support take-up.
The actual detailed CAD specimen has surface features: nearest center-upper and side-upper distances are approximately 1.2184 mm and 5.0891 mm. This animation uses the nominal rectangular envelope and configured clearances above. Actual contact detection and compliance correction must come from the qualified test method.
After the reverse phase, force must be released and the specimen confirmed re-seated on both lower supports before the empty fork can re-enter. The result/re-seating state models that confirmation; a physical sensor and method must qualify it.
Torsion through the same Instron load cell
The common shuttle moves the bending fixture out and the torsion fixture in. The upper center roller presses the torsion moment arm; the lower center cassette stays retracted. Both pneumatic end clamps retain the specimen, and one shaft rotates while the opposite end remains fixed.
θ = atan(δlever / a) [radians]
a = 64.95 mm; δlever = 2.2 mm; F = 11 N
M = 0.71445 N m; θ ≈ 1.9400°
The proposed kinematic model uses a fixed vertical force line and a sliding/rolling contact along the lever: a is the horizontal perpendicular distance to the shaft axis. Therefore δ = a tan θ. A fixed material-point follower would require a different geometric relation. F × a is the nominal moment estimated from axial force; actual conversion must account for angle-dependent contact reactions and effective leverage. Subtract contact pickup, frame compliance and backlash before deriving angle; calibrate the effective arm and angle against the real fixture. Roller radius, contact migration, parasitic bending and shaft compliance are not calibrated by this animation. A rotary encoder is the preferred direct-angle option.
Displayed channels: torque–time, torque–angle, angle–time, raw head displacement–time and force–time. Pneumatic pressure proves clamp supply; it is neither specimen force nor torque. No torsional shear-stress claim is made without an approved product section model.
Before mode indexing, verify the empty cell, retracted lower center pin and head at TRANSFER +220 mm. Lock the shuttle at its new station, then set BEND +150 mm or TORSION 0 mm. SERVICE +214 mm is a separate head position used only at the locked tooling service station.
Real Instron integration
Instron documents a Bluehill Universal API for metadata transfer, automation actions and results integration. Its official API overview describes a .NET/WCF integration route. Confirm the installed machine model/serial number, controller, Bluehill version, licensed modules, supported SDK and method permissions with Instron before implementation. The browser does not call undocumented machine endpoints.
| Contract | Required behavior |
|---|---|
| Job request | Unique request ID, specimen/lot/tray/pocket ID, method and revision, product dimensions, fixture/tool IDs and calibration references. |
| Acknowledgement | Match the same ID through accepted → running → complete/error. A timeout does not trigger an automatic second start. |
| Raw trace | Preserve timestamp or sample index, calibrated force (N), head displacement (mm), units, zero convention and sampling rate. Retain the original OEM export. |
| Completion | Wait for a completed result/export; verify file completion, job identity, monotonic time, expected columns, finite values and row count before marking the record complete. |
| Live display | Use OEM-supported event/status updates. Real live waveform access is subject to the installed API; if unavailable, display progress during the test and import the full waveform after completion. |
| Derived channels | Retain raw channels unchanged. Save dimension/arm inputs, gap/compliance calibration, units and calculation revision beside each derived trace. |
| Fault or disconnect | Stop issuing requests, preserve identity and partial data, quarantine the specimen where required and reconcile the actual Instron job before retry. Never convert missing data into a pass. |
| Network boundary | Run the adapter on the approved on-site Windows/OT network. Use authenticated outbound result transfer to the dashboard. Machine safety remains in validated hardware/PLC functions. |
Bidirectional ramps, holds and cyclic sequences must be implemented in an approved Instron method. See Bluehill Universal and TestProfiler for the official software scope; this simulator is not an Instron method file or an executable robot program.
Download the measurement and integration PDF
Every specimen keeps its evidence
The demonstration samples at 50 Hz: 1,201 samples for the 24-second bending method and 901 samples for the 18-second torsion method. Playback speed changes the animation pace, not the stored time base or sample count. Each record contains its ID, mode, robot, product, tray/pocket ownership, timestamps, settings, raw/derived channels and completion or quarantine status.
Completed traces are saved to the site's database after acknowledgement and can be selected again after reloading in the same browser session. The session cookie identifies these private demo records; clearing it removes that browser's access. CSV exports contain all samples, and JSON exports retain the full record metadata. A storage error remains visible with retry and export controls. No physical acceptance result is assigned.
“Complete demo” uses eight bending specimens, explicitly replaces the input/output trays, then tests eight fresh torsion specimens. It does not silently reuse a potentially damaged board in another destructive test.
Remaining engineering release inputs
Customer-approved contact/keep-out zones and allowable strain; bidirectional load capacity of the center cassette and lock; sensor diagnostic coverage; tolerance and continuous cable/hose sweep; ABB controller/RobotStudio path qualification; installed Instron travel, load-cell range and compression/tension calibration; metrology correlation; ESD/cleanroom acceptance; failure recovery and data-retention policy. Existing selected service clearances include a 0.681 mm gap and are not released manufacturing margins.
Formula references: Instron flexural-test guide · UNSW engineering formula sheet. Sources explain the method basis; the numeric forces here remain artificial.