Solenoid and Electromagnet Pull Force, Coil Design and Heating

Design a plunger solenoid, C-core or pot electromagnet. Compare force versus air gap with nonlinear steel, fringing, coil winding and cyclic heating.

#co-energy #coil design #cyclic heating #electromagnet #FEM verification #nonlinear B–H #Roters fringing #solenoid

This solenoid force calculator helps size flat-ended plungers, C-core actuators and pot-core holding electromagnets for locks, valves, relays and fixtures. It connects the magnetic circuit to the winding and its temperature, so a stronger cold pull does not hide the loss of current after the coil warms.

Choose the actuator geometry and enter its pole size, winding window and frame dimensions. For round designs, pole size is the central diameter. For a C-core it is the width of each rectangular pole, with stack depth entered separately. The C-core window has twice the winding radial depth between its limbs. The plunger has a fixed stop halfway along the winding and a sliding return sleeve. Drag the blue handle on the moving steel to change the air gap. The dimension in the section includes the residual nonmagnetic contact film.

The coil-design mode estimates turns from the insulated wire diameter, whole radial layers, whole turns per layer and winding utilization. Select an AWG copper wire and supply voltage, then compare alternative gauges in the winding table. If you already have a coil, choose measured turns and resistance instead. Keep its actual gauge and window dimensions: they still determine copper mass and cooling area.

Read the hot and cold settled-current force curves together. The cold curve uses ambient-temperature resistance; the hot curve uses the highest winding temperature in a repeating duty cycle. Duty fraction and cycle duration both matter. A short pulse train can have a much smaller temperature swing than long pulses with the same duty fraction. The thermal model includes temperature-dependent resistance, effective surface heat transfer, radiation and thermal storage. Set the heat-transfer coefficient from a representative mounting or a temperature-rise test.

This electromagnet pull force calculator includes saturation through an AISI 1010 B–H curve, finite fringing permeances and an equivalent winding-window leakage branch. It calculates force from magnetic co-energy and reports differential inductance and the local L/R time constant. A separate conservative current-rise bound covers the complete nonlinear ramp and requested stroke. If an on pulse is shorter, use an electrical transient model before relying on either force or heating. The force curves show settled-current capabilities, not travel times; points beyond the gap or B–H range are hidden. Springs, friction, gravity, valve pressure forces, motion-induced voltage and the current rise during a short pulse require a coupled dynamic analysis.

Four independent XFEMM field-solver cases test the nonlinear pot-magnet force across 0.1–1 mm gaps with disclosed dimensions and excitation. The published NAFSA VEM40 points illustrate how air gap and keeper thickness affect a real holding magnet. They apply to that manufacturer's hot 24 V coil at 35 °C ambient with flat St37 plates. They are a separate reference, not a fitted prediction for your geometry.

Use the tool for design comparisons and preliminary sizing. Conical tips, thin or perforated workpieces, severe leakage, local saturation, winding hot spots and complex heat paths need measurements or a field and thermal model. Pull force here is normal to aligned faces; it does not establish a permissible lifting load or a sliding capacity.

How it is calculated

The magnetic circuit has a working branch and a parallel, energy-equivalent window-leakage branch, with nonlinear shared steel. A bracketed root solve enforces ampere-turn balance using all 38 FEMM AISI 1010 B–H points (monotone cubic interpolation). Radial yoke and keeper MMF and energy are integrated with two-point Gaussian quadrature on geometrically spaced annular intervals (24 intervals by default). Inner-boundary flux densities are checked explicitly. C-core limbs, back and armature have their own areas.

Each closing face has gt=g+gresg_t=g+g_{res}. The direct gap permeance is μ0A/gt\mu_0 A/g_t. Roters flux tubes add a half-cylinder contribution 0.26μ0p0.26\mu_0 p and a finite half-ring contribution μ0pln⁡(1+2s/gt)/π\mu_0 p\ln(1+2s/g_t)/\pi per aligned edge perimeter pp. Extended keeper edges use half-length tubes. C-core corners include spherical quadrant tubes. Fringing extent is bounded by adjacent window space and steel thickness. These local perimeter approximations follow Piron (1999), printed pp. 29–33 (PDF 58–62), and Vrijsen (2016), Appendix A, printed pp. 181–184 (PDF 202–205), implementing the Roters method. A plunger also has the exact concentric-cylinder permeance of its radial return clearance and changing sleeve overlap.

For uniform winding excitation the effective leakage permeance is one third of the full-window transverse permeance: 2πμ0h/[3ln⁡(ri/rc)]2\pi\mu_0 h/[3\ln(r_i/r_c)] for round windows and μ0dh/(6w)\mu_0 d h/(6w) for the C-core. The one-third factor follows integration of the squared enclosed-turn fraction along the window; compare Piron pp. 25–26. Shared steel carries the equivalent total flux; the armature branch carries working flux.

Stored energy is the sum of air-gap, leakage and V∫0BH(b) dbV\int_0^B H(b)\,db contributions. Co-energy is W′=NIΦlinked−WW'=NI\Phi_{linked}-W. Positive closing force is −∂W′/∂g-\partial W'/\partial g at fixed current. Stationarity gives F=12Φworking2 dRair/dgF=\tfrac12\Phi_{working}^2\,d\mathcal R_{air}/dg, evaluated by a centred difference and independently checked by differentiating co-energy. In linear iron this reduces to 12I2dL/dx\tfrac12 I^2 dL/dx for closing displacement xx; differentiating the saturated secant inductance that way would be incorrect. The local electrical time constant uses Ld=dλ/dIL_d=d\lambda/dI divided by hot resistance. It does not bound a nonlinear current ramp. For the pulse applicability check, zero iron reluctance gives an upper bound Lmax=N2(Pleak+1/Rair,min)L_{max}=N^2(P_{leak}+1/\mathcal R_{air,min}) over the requested stroke and all currents. Positive incremental iron reluctance can only reduce this inductance. The conservative 99% rise-time bound is ln⁡(100)Lmax/R(Ta)\ln(100)L_{max}/R(T_a) for a voltage step from zero current; it can greatly exceed the actual rise time.

Turns are whole wire layers times whole turns per layer times utilization, rounded down. Resistance uses copper resistivity 1/581/58 Ω·mm²/m at 20 °C and a 0.00393 K⁻¹ coefficient. The repeating thermal cycle is bracketed directly, doubling the trial starting-temperature rise until the cycle's end is below its start. The supported temperature range ends at 300 °C. The shooting solution integrates CT˙=uV2/R(T)−A[h(T−Ta)+ϵσ(TK4−TaK4)]C\dot T=uV^2/R(T)-A[h(T-T_a)+\epsilon\sigma(T_K^4-T_{aK}^4)], with adaptive Dormand–Prince integration in each on/off subdivision. Thermal energy input, heat rejection and storage are integrated as separate states. Copper density 8960 kg/m³ and heat capacity 385 J/(kg·K) follow RSC's room-temperature data.

Assumptions

Flat, aligned faces; a complete C-core armature or a round keeper covering both pot poles. The plunger's stop is halfway along the coil, with a front return sleeve of overlap h/2+t−gth/2+t-g_t. Mean iron path lengths are fixed during the virtual displacement. Steel is isotropic and reversible: no remanence, hysteresis, eddy-current delay or temperature dependence of its B–H curve.

The leakage network is a lumped energy-equivalent approximation; it does not resolve leakage flux separately at every axial position. The winding uses a rectangular lattice with a user-selected enamel build and utilization; leads and terminals are excluded unless included in measured resistance. The AWG comparison always designs new windings in the stated window, even in measured-coil mode.

The coil and additional thermal mass share one temperature. The effective nonradiative coefficient includes mounting, bobbin and enclosure effects; radiation is added separately. Ambient and surrounding radiative temperature are equal. Current reaches its DC value during each on interval; off intervals assume rapid current extinction and put no magnetic turn-off energy back into the winding.

Where it applies

Use for preliminary design of flat-ended DC actuators. The low-carbon-steel curve is illustrative: machining, annealing and material batch can change it. The selected point and every hot and cold stroke point are checked against the final 2.471238 T B–H point. Above it, the law continues with vacuum incremental permeability to identify the excursion; selected results are extrapolated and a warning appears. Such stroke points are hidden. Total face gaps above half the smaller pole dimension are also hidden; reduce the selected gap if it exceeds that limit.

Round edges use locally planar Roters tubes. Curved-edge interactions, distributed leakage and local corner saturation need a field solver. Large gaps, very thin keepers, pole chamfers, misalignment and open-ended air-core solenoids fall outside the intended model. Radial quadrature resolves the radial magnetic-circuit integral, not a two-dimensional field. Refinement from 24 to 48 intervals is checked on a thin, saturated pot yoke and keeper, including the inner-boundary maximum B.

The force sweep solves 121 points by default, clustered near closure. Traces use solved values with gaps in unsupported regions. Refine stroke samples for a sharp saturation knee and halve the force-difference increment to check sensitivity. Doubling the thermal subdivisions is also checked. Coloured circuit paths represent section flux densities; radial paths use the computed flux divided by the local annular steel area. No spatial field is inferred. Winding dots are sampled for legibility; turns use every layer.

All forces are settled-current capabilities. The local differential L/R may be much smaller than the early-ramp time constant in saturating steel. A pulse shorter than the displayed conservative 99% rise-time bound needs a nonlinear electrical transient model to establish force and heating. Passing this bound does not predict motion, release time or flyback behaviour. The bound includes the closest point of the requested stroke, so it can be conservative for a magnet operated only at a larger gap.

Temperature is a winding-average estimate, not a hot-spot prediction. The single-temperature model stops at 300 °C and uses fixed room-temperature copper heat capacity. Choose heat transfer for the actual mounting. Off intervals assume rapid current extinction with no magnetic turn-off energy returned to the winding. Very short electrical pulses violate the settled-current heating approximation even when the periodic thermal solution exists.

The independent nonlinear pot-magnet benchmark uses the external XFEMM 4.0 field solver, weighted Maxwell stress, disclosed geometry, 1000 ampere-turns and the same 38 B–H input points. Four gaps from 0.1 to 1 mm test the complete saturation, fringing and leakage circuit without tuning its formulas. Its declared 20% comparison band is a preliminary-design acceptance criterion, not a guaranteed error bound for other geometries. The circuit underpredicts these four FEM forces by 9.6–16.2%; mesh refinement and a doubled air box change each FEM result by less than 0.1%. Geometry, excitation, material data and complete reproduction code are in attached source s8.

NAFSA VEM40 data illustrate a manufactured magnet; its undisclosed internal geometry prevents quantitative circuit validation. The Meeker case verifies the ideal two-gap limit; Multicomp checks cold electrical data. Neither the generated FEM benchmark nor these cases replaces prototype force and temperature measurements.

References

  • Roters, H. C., Electromagnetic Devices, 1st ed., Wiley, 1941. https://books.google.com/books?id=QRuaAAAAIAAJ . Flux-tube equations checked through the open theses below; original full text was not used.
  • Brauer, J. R., Magnetic Actuators and Sensors, 2nd ed., IEEE/Wiley, 2014. DOI: 10.1002/9781118779262 . Background reference for nonlinear magnetic circuits and co-energy.
  • Piron, R. (1999), University of Glasgow doctoral thesis, Chapter 2, printed pp. 20–33 and 62–63. https://theses.gla.ac.uk/1874/1/1999pironphd.pdf
  • Vrijsen, N. H. (2016), Eindhoven doctoral thesis, Appendix A, printed pp. 181–184. https://pure.tue.nl/ws/files/41301987/20161027_Vrijsen.pdf
  • FEMM material library, 1010 Steel: 38 B–H points. https://raw.githubusercontent.com/cenit/FEMM/master/bin/matlib.dat
  • Meeker, D., DC and AC Force of a Gapped EI-Core Inductor. https://www.femm.info/doku/doku.php?id=acforceexample
  • Multicomp, MCSMT-3257L12STD solenoid data sheet, 04 October 2011, pp. 1–2. https://www.farnell.com/datasheets/1490810.pdf
  • NAFSA, Holding Electromagnets, 2024, printed p. 102, Tables 1–2. https://nafsa.es/wp-content/uploads/2023/10/Serie_Holding-electromagnets-EN_1_2024.pdf
  • Royal Society of Chemistry, Copper: density and specific heat capacity. https://periodic-table.rsc.org/element/29/copper
  • Built-in wires.awg reference table: ASTM B258 diameter relation; annealed copper resistance from IEC 60028. No standard tables are reproduced.
  • XFEMM 4.0.0 (2023), independent axisymmetric nonlinear pot-magnet benchmark generated 2026-11-08; attached source s8 contains geometry, B–H input, generator, mesh-refinement results and provenance hashes. https://github.com/crobarcro/xfemm/releases/tag/v4.0

Inputs

InputSymbolUnitDefault
Actuator geometrygeometryFlat plunger solenoid
Air gapgmm2
Residual contact gap
Equivalent nonmagnetic film at each closing face, including roughness.
g_resmm0.05
Stroke curve endg_maxmm5
Pole size
Diameter for round cores; pole width for the C-core.
d_coremm12
Winding lengthh_coilmm24
Winding radial depthw_coilmm8
Frame thicknesst_framemm5
Armature thicknesst_armmm6
C-core stack depthdepthmm20
Radial clearanceclearancemm0.1
Enamel thickness per sideenamelmm0.015
Winding utilization
Fraction of the rectangular wire lattice actually wound.
utilization0.9
Coil datawindingDesign from window and AWG
Copper wire gaugewireAWG 28
Measured turnsN_set1385
Measured resistance at 20 °CR_setohm11.07
Supply voltageV_supplyV12
Duty fractionduty0.4
Cycle durationperiods20
Ambient temperatureT_amb°C25
Effective nonradiative coefficient
Include bobbin, enclosure and mounting effects in this effective coefficient.
h_effW/(m^2*K)12
Surface emissivityemissivity0.8
Additional thermal capacity
Heat capacity of bobbin and steel that follows the winding temperature.
C_addedJ/K0
Winding temperature limitT_limit°C130
Force differentiation incrementdgmm0.002
Stroke samplesn_stroke121
Steel modelsteel_modelAISI 1010, nonlinear
Roters edge fringingfringingtrue
Distributed window leakageleakagetrue
Linear relative permeabilitymu_linear2500
Thermal subdivisions per phasen_thermal4
Radial quadrature intervalsn_radial24

Results

With the default inputs above, the calc gives:

ResultValue
Hot settled pull force10.049 N
Cold settled pull force13.954 N
Coil turns1,346
Hot resistance21.911 Ω
Hot current0.548 A
Hot on-state power6.572 W
Cycle-average power2.635 W
Peak coil temperature72.1 °C
Peak temperature rise47.1 K
Differential inductance313.4 mH
Local L/R time constant14.301 ms
Maximum steel flux density1.138 T
Temperature comparisonBelow winding limit
99% rise-time upper bound1.002 s

Checked against sources

Each case below was calculated by this calc and agrees with the value from its source (within the tolerance where one is shown).

Published ideal two-gap force

Source: Meeker, DC and AC Force of a Gapped EI-Core Inductor, Circuit Theory Force: 66 turns, 1 A DC, central pole 0.5 in² and two 0.025 in gaps; F=1.09478 N. The E-core outer poles together have the central pole area, so this is the equal-area two-gap C-core equivalent with ideal iron and no fringing/leakage. Its inductance is 1.39 mH (Meeker, Inductance Calculation Example). https://www.femm.info/doku/doku.php?id=acforceexample ; https://www.femm.info/doku/doku.php?id=InductanceExample

Inputs: Actuator geometry = C-core and armature; Steel model = Infinite permeability benchmark; Roters edge fringing = no; Distributed window leakage = no; Coil data = Measured turns and resistance; Measured turns = 66; Measured resistance at 20 °C = 1 Ω; Supply voltage = 1 V; Ambient temperature = 20 °C; Duty fraction = 0; Air gap = 0.634 mm; Residual contact gap = 0.001 mm; Pole size = 12.7 mm; C-core stack depth = 25.4 mm; Force differentiation increment = 2 × 10⁻⁴ mm; the others as above.

ResultSourceThis calc
Hot settled pull force1.095 N1.095 N
Secant inductance0.00139 ± 2 × 10⁻⁶ H0.00139 H

Multicomp published cold coil

Source: Multicomp MCSMT-3257L12STD, 04/10/11 V1.1, PDF p. 2: R20=11.07 Ω, 1385 turns, 12 V continuous; rated 13 W and 1500 ampere-turns at 20 °C. Rounding explains 13.008 W and 1501.355 At from the printed coil values. This validates the electrical calculation, not this product's undisclosed internal magnetic geometry. https://www.farnell.com/datasheets/1490810.pdf

Inputs: Coil data = Measured turns and resistance; Ambient temperature = 20 °C; Duty fraction = 0; the others as above.

ResultSourceThis calc
Cold on-state power13 ± 0.02 W13.008 W
Hot magnetomotive force1,500 ± 3 A1,501 A

Published 1010 steel point

Source: FEMM material library, 1010 Steel: B=1.602049 T at H=3168.036204 A/m. Verifies the material law and units independently of the actuator dimensions. https://raw.githubusercontent.com/cenit/FEMM/master/bin/matlib.dat

Inputs: the default inputs above.

ResultSourceThis calc
Material reference field strength3,168 A/m3,168 A/m

NAFSA hot force-curve transfer

Source: NAFSA Holding Electromagnets (2024), p. 102 Table 2, VEM40 with 6 mm St37 plate: 400,245,160,30,5 N at gaps 0,0.1,0.2,0.5,1 mm, hot 24 V winding and 35 °C ambient. Data-transfer check only; the internal pole and winding dimensions are undisclosed, so no formula is adjusted to reproduce the product. https://nafsa.es/wp-content/uploads/2023/10/Serie_Holding-electromagnets-EN_1_2024.pdf

Inputs: the default inputs above.

ResultSourceThis calc
VEM40 with 6 mm plate [1]400 N400 N
VEM40 with 6 mm plate [2]245 N245 N
VEM40 with 6 mm plate [5]5 N5 N

Low-duty thermal shooting regression

Source: Independent reviewer reproduction: directly bracketed periodic solution for AWG20, 12 V, 0.1% duty, 20 s cycle, emissivity zero, h=1 W/(m²·K) gives 98.45 °C. This regression reproduces the review's reported ODE solution; it is not an external force benchmark.

Inputs: Copper wire gauge = AWG 20; Duty fraction = 0.001; Surface emissivity = 0; Effective nonradiative coefficient = 1 W/(m²·K); the others as above.

ResultSourceThis calc
Peak coil temperature98.45 ± 0.1 °C98.446 °C

Nonlinear XFEMM force at 0.1 mm

Source: External XFEMM 4.0.0 nonlinear axisymmetric field solution generated for this calculator, attached s8 pp. 1–3: 0.1 mm total face gap, 1000 A-turn, 12 mm pole, 24×8 mm winding window, 5 mm back/shell, 6 mm keeper, 0.1 mm radial clearance, all 38 FEMM 1010 B–H points. Weighted Maxwell-stress force 186.414160 N. Coarse-to-fine mesh and doubled air-box tests change each reference by <0.1%. The 20% reduced-circuit acceptance band covers uniform-section flux, lumped leakage and local Roters tubes; it is not FEM discretization error or a guaranteed design accuracy. No circuit formula was fitted. This is an independent numerical method, not published measurement. Software: https://github.com/crobarcro/xfemm/releases/tag/v4.0

Inputs: Actuator geometry = Pot-core holding magnet; Coil data = Measured turns and resistance; Measured turns = 1,000; Measured resistance at 20 °C = 1 Ω; Supply voltage = 1 V; Ambient temperature = 20 °C; Duty fraction = 0; Residual contact gap = 0.001 mm; Air gap = 0.099 mm; the others as above.

ResultSourceThis calc
Hot settled pull force186.4 ± 37.283 N165.2 N
Hot current1 A1 A

Nonlinear XFEMM force at 0.2 mm

Source: External XFEMM 4.0.0 nonlinear axisymmetric field solution generated for this calculator, attached s8 pp. 1–3: 0.2 mm total face gap, 1000 A-turn, 12 mm pole, 24×8 mm winding window, 5 mm back/shell, 6 mm keeper, 0.1 mm radial clearance, all 38 FEMM 1010 B–H points. Weighted Maxwell-stress force 154.405697 N. Coarse-to-fine mesh and doubled air-box tests change each reference by <0.1%. The 20% reduced-circuit acceptance band covers uniform-section flux, lumped leakage and local Roters tubes; it is not FEM discretization error or a guaranteed design accuracy. No circuit formula was fitted. This is an independent numerical method, not published measurement. Software: https://github.com/crobarcro/xfemm/releases/tag/v4.0

Inputs: Actuator geometry = Pot-core holding magnet; Coil data = Measured turns and resistance; Measured turns = 1,000; Measured resistance at 20 °C = 1 Ω; Supply voltage = 1 V; Ambient temperature = 20 °C; Duty fraction = 0; Residual contact gap = 0.001 mm; Air gap = 0.199 mm; the others as above.

ResultSourceThis calc
Hot settled pull force154.4 ± 30.881 N130.4 N
Hot current1 A1 A

Nonlinear XFEMM force at 0.5 mm

Source: External XFEMM 4.0.0 nonlinear axisymmetric field solution generated for this calculator, attached s8 pp. 1–3: 0.5 mm total face gap, 1000 A-turn, 12 mm pole, 24×8 mm winding window, 5 mm back/shell, 6 mm keeper, 0.1 mm radial clearance, all 38 FEMM 1010 B–H points. Weighted Maxwell-stress force 91.906277 N. Coarse-to-fine mesh and doubled air-box tests change each reference by <0.1%. The 20% reduced-circuit acceptance band covers uniform-section flux, lumped leakage and local Roters tubes; it is not FEM discretization error or a guaranteed design accuracy. No circuit formula was fitted. This is an independent numerical method, not published measurement. Software: https://github.com/crobarcro/xfemm/releases/tag/v4.0

Inputs: Actuator geometry = Pot-core holding magnet; Coil data = Measured turns and resistance; Measured turns = 1,000; Measured resistance at 20 °C = 1 Ω; Supply voltage = 1 V; Ambient temperature = 20 °C; Duty fraction = 0; Residual contact gap = 0.001 mm; Air gap = 0.499 mm; the others as above.

ResultSourceThis calc
Hot settled pull force91.906 ± 18.381 N76.994 N
Hot current1 A1 A

Nonlinear XFEMM force at 1 mm

Source: External XFEMM 4.0.0 nonlinear axisymmetric field solution generated for this calculator, attached s8 pp. 1–3: 1 mm total face gap, 1000 A-turn, 12 mm pole, 24×8 mm winding window, 5 mm back/shell, 6 mm keeper, 0.1 mm radial clearance, all 38 FEMM 1010 B–H points. Weighted Maxwell-stress force 45.039733 N. Coarse-to-fine mesh and doubled air-box tests change each reference by <0.1%. The 20% reduced-circuit acceptance band covers uniform-section flux, lumped leakage and local Roters tubes; it is not FEM discretization error or a guaranteed design accuracy. No circuit formula was fitted. This is an independent numerical method, not published measurement. Software: https://github.com/crobarcro/xfemm/releases/tag/v4.0

Inputs: Actuator geometry = Pot-core holding magnet; Coil data = Measured turns and resistance; Measured turns = 1,000; Measured resistance at 20 °C = 1 Ω; Supply voltage = 1 V; Ambient temperature = 20 °C; Duty fraction = 0; Residual contact gap = 0.001 mm; Air gap = 0.999 mm; the others as above.

ResultSourceThis calc
Hot settled pull force45.04 ± 9.008 N40.731 N
Hot current1 A1 A

Related Resources

Responses

Your email address will not be published. Required fields are marked *

L ading...