FrameScore

Preview

Get Started

Create Your Own Frame

Loading

Orientation If the frame shows on its side, upside down or the wrong way round, turn it here.
Suspension Pickups
Reset

–/100

    –
    Torsional stiffness axle to axle, Nm/deg (estimate)
    –
    Frame Mass
    –
    Nm/deg per kg
    –
    Bending stiffness, N/mm
    –
    Of target weight
    –
    Gussets resolved

    Torsional stiffness, front and rear

    Real-world stiffness estimate

    Calibrate to your FEA or rig test

    Have an FEA result or a rig test for this frame? Enter it and FrameScore scales every stiffness estimate on this device to match it, for this frame and the next ones you design. For a fair match, the FEA should hold the rear suspension pickups and twist the front pickups, using the same pickups as Edit Pickups.

     

    Biggest gains

      Gussets

      Joints are checked for tight angles, offset centrelines, weak pickups, crossed tubes and unbraced corners, against FIA Appendix J Art. 253-8 and structural welding guidance. Tap a joint to see the gusset on the model.

        Where It Sits Nm/deg

          Bare frames and complete cars aren't directly comparable: panels, glass and subframes add a lot of stiffness to a finished car.

          Your reference frames

          Save frames you know work. Each new design is compared against them, and any measured stiffness you enter calibrates the estimates. Saved on this device.

            Notes

              Use your own frame

              Upload the frame as a STEP file (.step or .stp) exported from SolidWorks, Fusion, Rhino, Alias or CATIA. FrameScore finds every round tube from its cylindrical faces, reads diameter and wall, follows bends, and rebuilds the nodes where coped tubes meet. Brackets, plates and holes are ignored.

              After loading, use Edit pickups and tap the nodes where the wishbones, pushrods and dampers mount. FrameScore sorts them into front and rear, left and right, and remembers them for this frame on this device. A 3D DXF of tube centrelines, or the frame file below, also works.

              Not sure how to set up your STEP file? Download the example, open it in your CAD to see how it's built, then set your frame up the same way.

              Download example STEP file Sample roadster frame: 79 round tubes including a bent roll hoop, in millimetres (11.8 MB).

              • Bare chassis only: every tube as its own solid body. Leave out suspension parts, panels and the engine; brackets, tabs and gussets are ignored anyway.
              • Real round tube: hollow tube with the true outer diameter and wall, so FrameScore can read both. Square and rectangular tube aren't read yet.
              • Ends: coped (fish-mouthed) or square-cut both work; FrameScore extends the centrelines to find each joint. Bent tubes are read as one tube.
              • Units and axes: millimetres, Z up and X along the car, as in the example. If yours differ, change Units in file and Orientation above before uploading.
              • Export: STEP AP214 or AP242 (.step or .stp).

              To check it works, upload the example first: it should read 75 tubes and 48 nodes.

              No CAD? You can describe the frame in a spreadsheet instead. Number every joint (node). With the frame on a level floor, measure each node's X along the car, Y across it (left negative, right positive) and Z up from the floor; a tape, a plumb bob and a string line along the centreline are enough. Then list each tube by the two nodes it joins, with its outer diameter and wall. Save it as a .csv and upload it, one row per line:

              UNITS,mm
              NODE,F1,0,-230,180
              NODE,F2,0,230,180
              TUBE,F1,F2,25.4,1.6
              PICKUP,FL,F1

              The box below shows the current frame in this format: copy it into a spreadsheet as a starting point.

              Suspension pickups

              How the score works

              The design score checks the frame against the layout rules that proven space frames follow. Each check looks at geometry only:

              • Torsional stiffness (25%): 70% from the whole-chassis stiffness, axle to axle, against published figures for the class you pick (Seward recommends at least 1,000 Nm/deg for any racing car), and 30% from the balance between the front and rear halves.
              • Triangulation (20%): share of flat bays braced by a diagonal. Bays you keep open for the driver, engine or drivetrain don't count against you.
              • Suspension load paths (15%): every pickup you designate should sit on a node with tubes running in three directions.
              • Chassis weight (10%): frame mass as a share of your target vehicle weight. For steel space-frame kit and track cars, 12–17% is the desired range; under 9% is underweight and over 20% unnecessarily heavy. For cars under 450 kg the range is 10–20%.
              • Clean nodes (10%): from the STEP geometry, how many coped joints have centrelines meeting at one point.
              • Joint angles and gussets (10%): joints still needing a gusset or a decision from you.
              • Diagonal angles (5%): triangles with an angle under 20°.
              • Unbraced tube lengths (5%): tubes longer than 35× their diameter between nodes.

              Real-world stiffness. The beam model assumes perfectly rigid joints, so it reads high. Real welds, fish-mouth gaps and joint flex typically cost 15–30% of torsional stiffness. FrameScore shows the theoretical value and an estimated real-world value, using either a fixed reduction you choose (15–30%) or a node-by-node assessment of every joint: wall slenderness, tube size ratio, joint angle, number of tubes and, from STEP files, centreline offsets, weighted by the twist load through each joint. The design score is shown both ways. A measured frame saved as a reference replaces the correction.

              Front and rear stiffness. One twist test is run: rear pickups held, front pickups twisted. FrameScore measures the twist at every cross-section along the frame, takes the frame between the innermost front pickup and the innermost rear pickup, splits it at the middle, and divides the torque by the twist across each half. The loaded and held pickup regions are left out because they are forced rather than free; if the split can't be measured reliably (for example when pickups reach far along the car), only the whole-chassis value is scored. The two halves act like springs in series, so the whole chassis is always softer than either half and the weaker end sets it; that's why the average of the halves is shown but never scored. The balance score is 100 up to a 1.5:1 ratio, 70 at 2:1, 30 at 3:1 and 0 at 4:1 or worse. A ratio of 2:1 or more is flagged as a significant imbalance. These limits are working rules of thumb.

              Gusset rules. A gusset is suggested where two tubes meet at under 30° (AWS D1.1 and AISC Design Guide 24 treat 30° as the minimum branch angle for a full-strength weld), where centrelines miss by more than a quarter of the tube diameter, where a pickup node is braced in one plane, where two tubes cross (FIA Art. 253-8.3.2.1.5 requires at least two gussets where cage diagonals or door bars cross), and at corners where only two tubes meet. Suggested sizes follow FIA Art. 253-8.2.14: U-shaped sheet at least 1.0 mm thick, with legs 2–4× the largest tube diameter from the corner. FIA rules apply to safety cages; elsewhere treat them as good practice.

              Weight benchmarks. A GT40-replica builder's 150 kg frame for a 1,000 kg car is 15%; another builder's 375 lb frame for a sub-2,400 lb car is about 16%. Formula Student steel frames run from about 11% to 15% of wet car weight. Mass shown is tube mass plus gussets you mark as added; brackets, tabs and weld are extra.

              Bent tubes are kept as one continuous tube. Bends are checked against the FIA minimum centreline radius of 3× the tube diameter (Art. 253-8.3.3).

              The stiffness figure is a quick estimate, worked out automatically with no meshing, loads or setup. It treats every tube as a beam and every joint as fully welded, so it reads higher than a rig test. Save one of your measured frames as a reference with its measured value and every estimate is calibrated from then on. It's a screening number for comparing designs, not a replacement for your FEA.

              The rule thresholds (20°, 30°, 35× diameter, a quarter-diameter offset), the weight bands and the kit car stiffness bands above the 2,000 Nm/deg minimum are working rules of thumb, to be tuned with your own experience.

              Biggest gains tries a 19.05 × 1.6 mm diagonal across every open, flat bay, recalculates each one, and ranks them by stiffness added.

              Sources: Derek Seward, Race Car Design, ch. 2; FIA Appendix J Art. 253-8; Steel Tube Institute on 30° branch angles; GT40s chassis weight thread; Berkeley FSAE car data. Stiffness: MDPI World Electric Vehicle Journal, Formula Student torsion study, GT40s forum, Autosport technical forum, DSport.