Flow simulation¶

Use flow simulation to see how air or water moves around a part, or how a spinning part moves the fluid. You can compare designs, check which shroud has less drag, see which way a fan pumps, or inspect the pressure on a blade.

Wind tunnel¶

In a wind-tunnel study, the part stays still while a stream flows past it.

1. Open a part, select its body and start Flow from the toolbar. The toolbar shows this operation only when the body is selected. You can also find it in the F1 palette regardless of your selection.
2. Leave the mode on Tunnel.
3. Choose the fluid: air or water. N3XD sets the default speed to 10 m/s for air and 1 m/s for water. Because water is a thousand times denser than air, this method cannot answer questions about water flowing at air-like speeds. A study set up through the AI assistant can use another fluid. The field then shows Custom with its density and viscosity, and keeps it until you choose air or water.
4. Set the direction the stream blows from and its speed.
5. Optionally choose a torque axis to see how strongly the stream twists the part around it, and in which direction. Select it the same way as a rotation axis (below). If you leave the field empty, the result omits this value. You can change the axis later without running the simulation again.
6. Check the flow domain, shown as a box around the part. N3XD sizes it automatically. To change it, drag a face or enter values in the fields; select Auto to restore the automatic size. A box that is too large to solve keeps its fields, so you can shrink it again.
7. Choose a resolution. Draft, standard and fine set approximate time budgets of one, five and ten minutes rather than mesh sizes. The line below shows the cell size and estimated runtime.
8. Select Solve flow. A ring in the corner notification fills as the solver works through its budget.

Rotation¶

Use rotation mode to simulate a part spinning around an axis, such as a fan, impeller, stirrer or turbine.

1. Switch the mode to Rotation.
2. Set the speed in rpm. Use a negative value to reverse the direction, and choose how many revolutions to simulate.
3. Choose the rotation axis by clicking the field and selecting an origin axis, a body edge or a circular edge such as a hub bore. N3XD uses the selected edge's axis of rotation and shows it as a dashed line. If the blades are a circular pattern, N3XD fills in the axis automatically.
4. Set the domain and resolution, then solve the study as described above. For a rotating part, the domain is a loose duct rather than a tunnel.

Reading the results¶

Results appear in the card on the right, and the corresponding fields are drawn in the viewport.

• Drag force or Thrust, and Moment: these show the force the fluid applies to the body and the moment caused by that force. A tunnel study calls the force drag, while a rotation study calls it thrust, because it is the force the fan pushes back against its mounting. Hover over the moment to see its components and reference point: the bounding-box centre for a tunnel, or the rotation axis for a rotation study.
• Torque @ axis (tunnel, when you selected a torque axis): how strongly the stream twists the part around that axis. Use it to check whether a vane, flap or unpowered rotor would turn, and in which direction. Positive values follow the right-hand rule around the axis. This result is as reliable as the drag force; the accuracy caveat below concerns parts you drive, not parts pushed by the flow.
• Flow rate (rotation): the amount of fluid the part moves, in litres per minute.
• Torque (rotation): an approximate estimate of how hard the motor must turn the part. Multiply it by the angular speed to estimate shaft power; see the accuracy notes below.
• Lattice: how many cells and steps the run used.
• converged or step cap (tunnel): whether the flow settled before the solver reached its step limit. A step-cap result is averaged over the final part of the run, but it is less reliable. Rotation studies instead show fixed run because they simulate exactly the number of revolutions you requested and do not use a settling test.

You can toggle three result layers: surface pressure on the body; streamlines, coloured by speed; and a cut plane through the domain that shows speed or pressure. Streamlines animate from the inlet to the outlet when the result appears, making the flow direction clear. The cut plane stays hidden until you turn it on.

If the solver detects a potential issue, it shows an amber message under the results. Messages explain the issue in plain language, such as a body reaching the damped edge of the domain, features too thin for the lattice to resolve, or a run that did not settle.

Accuracy¶

The solver represents the part with a grid of cubic cells. Its surfaces are therefore stepped rather than smooth, which makes the simulation fast but limits its accuracy.

• Use a finer resolution for thin features. At a coarse preset, a 3 mm fan blade may be only one or two cells thick, which is not enough to show how it affects the flow. The solver warns when the whole body is this thin; choose a finer resolution to improve the result.
• Use the results to compare designs and flow directions. The solver can reliably show which of two shrouds has less drag, but it cannot give a precise drag value to two significant figures.
• Treat a rotating body's torque as an order-of-magnitude estimate. For the fan used during development, three resolutions with cell sizes spanning a factor of 1.75 agreed on the torque direction and stayed within a factor of 1.3 in magnitude. This is enough to estimate "about six watts" of shaft power, but not enough to choose a motor. Flow rate and direction from the same run have similar accuracy. By contrast, Torque @ axis in a tunnel study measures a stationary body and is as reliable as the drag force shown beside it.
• Check the convergence badge. A step cap result did not settle. For a tunnel study, raise the flow-through count or make the domain coarser to give the solver more time. Rotation studies show fixed run instead; that is not a warning, and flow_throughs does not apply to them.