BPSIM

Windkessel model of the human aorta
SIMULATION 00:00
ARTaortic pressure Pc, mmHg
ECGsinus rhythm
FLOWejection into the aorta, ml/s— backflow
SYS --- / DIA --- / MAP ---mmHg, trend last 100 s

Conditions

Illustrative parameter sets for teaching, not clinical reference values.

Parameters

Heart

Arteries

Aortic valve

Model details equations, assumptions and constants
Ventricle E(t) Q(t) Aorta C Periphery Rp Zc P(t) Pc
L·dQ/dt = PLV − Pc − Zc·Q − K·Q|Q|
C(Pc)·dPc/dt = Q − Pc / Rp    P = Pc + Zc·Q

Left ventricle

Time-varying elastance [5, 6]: PLV = e(t)·Emax·(V − V0) + (1 − e(t))·Ppassive(V) − RV·e(t)·Q. The activation e(t) is a double-Hill curve [6] scaled by the systole duration. The passive filling curve is Ppassive = 0.26·(eB·(V − 15 ml) − 1) mmHg; its stiffness B is adjustable (default 0.03; lower B: compliant, dilated ventricle; higher B: stiff, hypertrophied ventricle). The ventricle fills from the filling pressure through the mitral valve. Stroke volume, ejection time and flow are therefore results of contractility, filling, valve and afterload.

Aortic valve and blood inertia

The valve gradient grows with the square of flow, ΔP = K·Q². By Bernoulli, ΔP = ½ρv² ≈ 4v² (v in m/s) [8], so K corresponds to a valve area of about √(4/K) cm² (K in 10−4 mmHg·s²/ml²). The blood in the outflow tract has an inertia L = ρ·l/A, from the pressure needed to accelerate the blood column, ΔP = ρ·l·dv/dt with v = Q/A (an inertance term as in the four-element windkessel [4], but placed in series here). With ρ = 1.06 g/ml, l = 2 cm and a diameter of 2 cm, L ≈ 0.0005 mmHg·s²/ml. Because L is small, ventricular and aortic root pressure differ only by L·dQ/dt and the valve gradient during ejection. Late in systole the flow continues briefly after the ventricular pressure has fallen below the aortic pressure, then reverses and closes the valve. This short backflow produces the dicrotic notch. In aortic regurgitation the closed valve leaks through a regurgitant orifice: blood flows back from the aorta into the ventricle throughout diastole, with the same Bernoulli relation. The ejection fraction is computed from the total ejected volume; with regurgitation the net forward stroke volume is smaller.

AV gradient readout

The “AV GRADIENT” block shows the pressure drop across the aortic valve, from the left ventricle to the aortic root, during ejection. In each time step the model computes it from the valve flow Q as ΔP = K·Q². This is the valve part only: the pressure needed to accelerate the blood (L·dQ/dt) and the drop across Zc are not included, so it corresponds to the gradient measured by Doppler echocardiography [8].

  • Peak: the largest instantaneous gradient of the last beat (peak instantaneous gradient). It is larger than the “peak-to-peak” gradient of a catheter pull-back, which compares the two pressure maxima at different times.
  • Mean: the average gradient over the time of forward flow.
  • vmax: the peak jet velocity derived from the peak gradient with the simplified Bernoulli equation, vmax = √(ΔPpeak/4) in m/s [8].
  • Valve area (under the valve resistance slider): A = √(4/K) cm² for K in 10−4 mmHg·s²/ml², from ΔP = 4v² and v = Q/A.
  • Regurgitation (only with a regurgitant orifice): backflow volume per beat and its share of the ejected volume (regurgitant fraction).

For orientation, guidelines grade aortic stenosis as severe at a mean gradient of at least 40 mmHg, vmax of at least 4 m/s or a valve area below 1.0 cm², and aortic regurgitation as severe at a regurgitant volume of at least 60 ml or a regurgitant fraction of at least 50 % [16, 17]. A normal valve in this model gives a peak gradient of a few mmHg and vmax below 1 m/s; the “Aortic stenosis” condition gives about 72/44 mmHg and vmax about 4.2 m/s.

Arteries

Three-element windkessel [2, 3]: the compliant aorta stores part of the stroke volume and drains through the peripheral resistance Rp = R − Zc, so pressure falls exponentially in diastole with τ = Rp·C. The characteristic impedance Zc of the proximal aorta adds a pressure component proportional to flow. Its default of 0.04 mmHg·s/ml lies in the range measured in adults without heart disease [9–11]; higher values are found in hypertension, especially isolated systolic hypertension [9, 12]. The difference between aortic root and Pc corresponds to the “excess pressure” of the reservoir–wave concept, which was found to be proportional to aortic flow [13]. As a guide, Zc and the pulse wave velocity are also calculated from the compliance: Bramwell–Hill, c = √(V / (ρ·C)) [7], for an aorta 50 cm long and 3.3 cm wide, and Zc = ρ·c/A. This guide value is approximate, because C stands for the whole arterial tree, of which the thoracic aorta holds only about half [13]. Compliance falls with pressure, C(P) = C0·e−0.012·(P − 100), with C0 set at 100 mmHg.

Display

The red trace and all pressure readouts show Pc; the dashed line shows the aortic root pressure P. In the flow chart, the optional dashed line shows the outflow at the end of the aorta into the periphery, Pc/Rp: it is nearly constant, because the compliant aorta turns the pulsatile ejection into a steady flow (windkessel function). The ECG is an illustration timed to the simulated beats and does not feed back into the model. With the heart model set to “Flow source”, ejection is imposed as in the original BPSIM [1]: a half-sine or skewed flow profile, constant compliance, and a notch produced by a brief change in compliance; the LV pressure outside ejection is then an illustration as well.

Fixed constants

ConstantValueUnitMeaningSource
Blood density ρ1.06g/mlInertance and Bernoulli gradientstandard value
Unloaded LV volume V015mlVolume at zero pressureassumption ([6] uses 20 ml)
Passive filling curveA = 0.26, B = slider (default 0.03)mmHg, 1/mlPpassive = A·(eB·(V−V0) − 1)assumption
Activation exponents1.32, 21.9–Double-Hill function e(t)[6]
Activation time constants0.75·Ts, 1.25·TssTs = systole duration slider; ratio 1.67 as in [6] (1.68); e(t) normalised to a peak of 1adapted from [6]
Internal resistance of the ventricle RV0.3mmHg·s/mlPLV reduced by RV·e(t)·Q during ejectionassumption
Mitral resistance0.005mmHg·s/mlFilling from the filling pressure, one-wayassumption
Outflow tractl = 2, d = 2cmInertance L = ρ·l/A ≈ 0.0005 mmHg·s²/ml, in seriesassumption; concept [4]
Aortic valve gradientΔP = K·Q²mmHgK = 4/A²·10−4 (A = valve area in cm²)[8]
Aortic regurgitationQ = −f(ΔP), R = 0.05mmHg·s/mlBackflow in diastole through the regurgitant orifice: K·Q² + R·Q = Pc − PLV, K = 4/AR²·10−4Bernoulli [8]; R assumed
Valve closure1 ml backflow, +0.01ml, mmHg·s²/ml²Closing volume and extra resistance of the closing leafletsassumption
Pressure dependence of compliance0.012, reference 100 mmHg1/mmHgC(P) = C0·e−0.012·(P−100) (elastance model only)assumption
Aorta for the PWV and Zc guidelength 50, diameter 3.3cmc = √(V/(ρ·C)), Zc = ρ·c/A[7]; dimensions assumed
Unit conversion1 mmHg = 1333.2dyn/cm²Converts ρ·l/A and ρ·c/A to mmHg·s²/ml and mmHg·s/mlphysical constant
Initial stateVLV 120, Pc 90 (flow model: aortic volume 125)ml, mmHg, mlStart values; the model reaches steady state within a few beatsassumption
Numerical integrationexplicit Euler, 1/3000sTime step of the simulation–
Flow model: ejection profilesx1.5(1−x)3.5; half sine; x1.2(1−x)1.2–Skewed (peak at 30 %), original BPSIM, aortic stenosis (peak at 50 %)[1] for the half sine; others assumed
Flow model: incisura+8 % of C for 0.1 s–Half-sine compliance change at the end of ejection (original: +0.1 ml/mmHg)adapted from [1]
Flow model: LV pressure outside ejectionτ 35 ms, 50 ms, 4 mmHg–Relaxation time constant, isovolumic contraction, atrial kick; filling from 3 mmHg towards the filling pressure within about 0.4 s (illustration)assumption
ECGQTc 0.40 ssShapes are illustrations timed to the simulated beats; QT scales with the RR interval (Bazett, QT = 0.40·√RR), PR and wave widths shorten moderately at high ratesassumption; Bazett formula

Adjustable parameters (sliders)

Parameter sets of the conditions

Values chosen for teaching (see “Model assumptions”). Emax was calibrated so that the elastance model reaches the stroke volume listed in the SV column; SV is set directly only in the flow model.

Orange: parameters that represent the primary cause. Blue: parameters that represent the physiological (regulatory or structural) compensation.

HR heart rate, Ts systole duration, R total peripheral resistance, C0 aortic compliance at 100 mmHg, Zc characteristic impedance, Emax contractility, Pfill filling pressure, B LV diastolic stiffness, K aortic valve resistance, AR regurgitant orifice area, SV stroke volume (calibration target).

Model assumptions without a specific source

  • Steepness of the pressure dependence of compliance (0.012 per mmHg).
  • Internal resistance of the ventricle RV = 0.3 mmHg·s/ml, mitral resistance 0.005 mmHg·s/ml.
  • Valve closure after 1 ml of backflow, with an extra closing resistance of 0.01 mmHg·s²/ml².
  • Passive filling curve and V0 = 15 ml; outflow tract and aortic dimensions.
  • All parameter values of the conditions, the ECG patterns and the scaling of activation time to systole duration.
Scientific sources
  1. Randall JE. Microcomputers and Physiological Simulation. Addison-Wesley; 1981. ISBN 0-201-06128-7. Original BPSIM model and parameters.
  2. Frank O. Die Grundform des arteriellen Pulses. Z Biol 1899;37:483–526. English translation: Sagawa K, Lie RK, Schaefer J. J Mol Cell Cardiol 1990;22:253–277. Windkessel concept.
  3. Westerhof N, Lankhaar JW, Westerhof BE. The arterial Windkessel. Med Biol Eng Comput 2009;47:131–141. Three-element windkessel, characteristic impedance.
  4. Stergiopulos N, Westerhof BE, Westerhof N. Total arterial inertance as the fourth element of the windkessel model. Am J Physiol 1999;276:H81–H88. Inertance in lumped arterial models. There, total arterial inertance is placed in parallel with Zc (dogs); this simulation instead uses a small outflow-tract inertance in series.
  5. Suga H, Sagawa K, Shoukas AA. Load independence of the instantaneous pressure-volume ratio of the canine left ventricle and effects of epinephrine and heart rate on the ratio. Circ Res 1973;32:314–322. Time-varying elastance.
  6. Stergiopulos N, Meister JJ, Westerhof N. Determinants of stroke volume and systolic and diastolic aortic pressure. Am J Physiol 1996;270:H2050–H2059. Double-Hill activation function with exponents 1.32 and 21.9 as in the paper (human set: Emax 2.31 mmHg/ml, Vd 20 ml, heart period 1 s). Here the time constants are scaled to the systole duration.
  7. Bramwell JC, Hill AV. The velocity of the pulse wave in man. Proc R Soc Lond B 1922;93:298–306. Pulse wave velocity and compliance.
  8. Hatle L, Angelsen B, Tromsdal A. Noninvasive assessment of aortic stenosis by Doppler ultrasound. Br Heart J 1980;43:284–292. Simplified Bernoulli equation, ΔP = 4v².
  9. Nichols WW, Conti CR, Walker WE, Milnor WR. Input impedance of the systemic circulation in man. Circ Res 1977;40:451–458. PubMed 856482 Zc in the ascending aorta usually 26–80 dyn·s·cm−5 without cardiovascular disease (mean 53), higher with coronary disease and hypertension.
  10. Murgo JP, Westerhof N, Giolma JP, Altobelli SA. Aortic input impedance in normal man: relationship to pressure wave forms. Circulation 1980;62:105–116. Zc 47 dyn·s·cm−5, PWV 6.7 m/s.
  11. Murgo JP, Westerhof N, Giolma JP, Altobelli SA. Effects of exercise on aortic input impedance and pressure wave forms in normal humans. Circ Res 1981;48:334–343. Zc unchanged during exercise (47 vs. 45 dyn·s·cm−5); mean pressure and pulse wave velocity rose, implying lower aortic compliance.
  12. Yano Y, Neeland IJ, Ayers C, Peshock R, Berry JD, Lloyd-Jones DM, Greenland P, Mitchell GF, Vongpatanasin W. Hemodynamic and mechanical properties of the proximal aorta in young and middle-aged adults with isolated systolic hypertension: the Dallas Heart Study. Hypertension 2017;70:158–165. PMC5518635 Zc about 70 dyn·s·cm−5 with optimal blood pressure and about 100 with isolated systolic hypertension.
  13. Wang JJ, O’Brien AB, Shrive NG, Parker KH, Tyberg JV. Time-domain representation of ventricular-arterial coupling as a windkessel and wave system. Am J Physiol Heart Circ Physiol 2003;284:H1358–H1368. Reservoir and excess pressure; excess pressure proportional to aortic flow; the thoracic aorta holds about 45 % of the windkessel volume (dogs).
  14. Avolio AP, Butlin M, Walsh A. Arterial blood pressure measurement and pulse wave analysis – their role in enhancing cardiovascular assessment. Physiol Meas 2010;31:R1–R47. Radial and central aortic waveforms (Figure 3), used to check the pressure shape.
  15. Yartsev A. Normal arterial line waveforms. Deranged Physiology. derangedphysiology.com Waveform components and dicrotic notch.
  16. Vahanian A, Beyersdorf F, Praz F, et al. 2021 ESC/EACTS Guidelines for the management of valvular heart disease. Eur Heart J 2022;43:561–632. Severity grading of aortic stenosis and regurgitation (not in the literature folder, not checked).
  17. Otto CM, Nishimura RA, Bonow RO, et al. 2020 ACC/AHA Guideline for the management of patients with valvular heart disease. Circulation 2021;143:e72–e227. Severity grading of aortic stenosis and regurgitation (not in the literature folder, not checked).

Program by Ralf Mrowka, © 1991–2026. Model parameters of the original BPSIM from James E. Randall, Microcomputers and Physiological Simulation, Addison Wesley, 1981, ISBN 0-201-06128-7. Educational simulation only. The values are model output and are not measurements of any person.