BiVACOR reference data

Before you use these numbers

Three figures do not reconcile

Found by an internal consistency check on 10 September 2026. These are unresolved. They have not been traced back to source, so each one may be a transcription error or a real property of the source document. Do not move any of them into params.py until that is settled.

Open findings

The left volute does not fit inside the stated envelope. The volute outer wall diameter is given as approximately 71 mm, claimed range 65 to 76 mm, from the patents. The whole device outer diameter is given as approximately 60 mm, from the magnetic bearing paper. Even the bottom of the claimed range exceeds the device diameter. The volute base circle, approximately 60 mm, equals the full device diameter on its own and leaves no wall. Likely explanation is that the patents describe a design family rather than the built device, or that the 2014 envelope predates the 2020 and 2023 filings, but the two are asserted here without reconciliation.
Rotor height and vane height share no stated datum. Rotor height is given as approximately 10 mm and right vane height as approximately 17.5 mm, so the vanes are taller than the rotor carrying them. Almost certainly rotor height means hub or disk thickness excluding vanes, but the source does not say so. Summing these for a stack height gives a wrong answer.
The 240 um working clearance sits below the patent minimum. Axial gaps are given as 250 to 350 um, typically 300 um, from the later patent. The working clearance is given as at least 240 um from the hemocompatibility paper. Probably design specification against operating condition, and defensible, but the 240 um figure is described below as the most transferable number in the set, and it undercuts the patent figure printed near it.

What did check out: central cavity diameter against right vane diameter is consistent with the stated radial gap, and both pressure and flow pairs are physically coherent, with higher head at lower flow on each side, which is the correct shape for a centrifugal pump curve. The patent and the bench paper agree with each other there rather than conflicting.

Tag	Meaning	Weight
[PAPER]	Peer reviewed journal article or reviewed conference paper	Highest
[REG]	Regulatory public document, FDA summary of safety and effectiveness	High, but device specific and dated
[PATENT]	Granted patent, BiVACOR Inc as assignee	Authoritative as disclosure, but states claimed ranges, not one built part
[TEAM-MAG]	Magazine article written by the BiVACOR engineers themselves	Reliable for round numbers, no method section
[MFR]	Manufacturer website	Marketing register, use only where nothing better exists

Part 1

The left ventricular pump

The closest thing to a published LVAD-scale centrifugal geometry from a device that has been implanted in a person. All values [PATENT-1][PATENT-2].

Left side of the BiVACOR rotor
Parameter	Value	Claimed range
Outer vane diameter	approx 50 mm	45 to 55 mm, or 48 to 52 mm; both filings list both brackets
Primary vane inner diameter	approx 25 to 30 mm	25 to 35 mm
Secondary vane inner diameter	approx 35 mm	30 to 40 mm
Vane height	approx 2 mm	1.5 to 3 mm
Primary vane outer thickness	approx 7.5 mm	6 to 15 mm
Vane inlet angle	approx 84 deg	82 to 86 deg
Vane outlet angle	approx 40 deg	30 to 50 deg
Volute base circle diameter	approx 60 mm	54 to 64 mm
Volute outer wall diameter	approx 71 mm	65 to 76 mm
Outlet throat area	approx 140 to 150 mm sq	60 to 250 mm sq
Outlet aspect ratio, width to height	approx 1.4 : 1	1:2 to 2:1
Cutwater angle	approx 45 deg	40 to 45 deg
Vane passage minimum flow area	min 150 mm sq	120 to 300 mm sq
Pressure rise at 6 L/min	approx 80 mmHg	60 to 100 mmHg
Axial sensitivity	200 um axial shift changes flow by at least 2 L/min, or head by at least 40 mmHg

Note the left vane height of about 2 mm against the right vane height of about 17.5 mm. That ratio is the whole trick of the device: the same rotation speed has to produce roughly 80 mmHg on the left and roughly 20 mmHg on the right, and the vane geometry, not the speed, carries that difference.

Real world LVADs, for envelope comparison, all rows [PAPER-9]
Device	Pump envelope	Mass	Prime or displacement	Speed	Flow	Suspension
HeartMate 3	69 mm dia x 30 mm high	200 g	50 cc prime	2000 to 5500 rpm	up to 13 L/min	Fully active electromagnetic, self bearing motor
HeartWare HVAD	50 mm dia	145 g	50 cc displacement	2400 to 3200 rpm	up to 10 L/min	Passive magnetic plus hydrodynamic thrust
HeartMate II	40 mm dia x 60 mm long	not stated	63 mL	not stated	3 to 10 L/min	Mechanical, blood immersed
CentriMag, external	not stated	not stated	31 mL	up to 5000 rpm	up to 9.9 L/min	Magnetic, gaps over 0.6 mm

The HeartMate 3 mass of 200 g in that source includes motor, inflow cannula, flexible recovery section, outflow graft, bend relief and all connecting hardware, so it is not a bare pump mass. Two further cautions. First, [PAPER-9] is a 2017 review describing "a recent design of the pump", which is a development configuration, not necessarily the approved commercial one. Second, its 13 L/min figure is higher than the up to 10 L/min quoted in the clinical literature [PAPER-11]. Where the two disagree, use 10 L/min and say which source you took it from. The FDA approval document for HeartMate 3, PMA P160054, approved 23 August 2017, contains a device description but gives no dimensions or masses at all [REG-2]. Do not cite it for geometry.

HeartMate 3 hemocompatibility design targets [PAPER-10]

CFD residence time, secondary flow path 27 to 798 min, main flow path 118 to 587 min.
Shear stress exposure above 150 Pa, HeartMate 3 against HeartMate II: 3.3 against 11 within pump volume, 134 against 604 on surfaces. The abstract renders both as "mm", which cannot be right for a volume and an area. Do not quote these two numbers without the full text in hand.
In vitro hemolysis, plasma free haemoglobin at 6 h, at 2, 5 and 10 L/min: 58, 74, 157 mg/dL, against 112, 123, 353 mg/dL for HeartMate II.
Eight 60 day bovine implants, average flow 5.6 to 6.4 L/min, no device failures, no thrombosis, no hemolysis.

Part 2

The complete artificial heart

Whole device envelope
Parameter	Value	Source
Outer diameter	approx 60 mm	[PAPER-1]
Height	approx 70 mm	[PAPER-1]
Mass	approx 650 g, described as slightly heavier than an adult human heart	[TEAM-MAG-1]
Housing material	Titanium	[TEAM-MAG-1][MFR-1]
Patient size limit	Body surface area over 1.4 m sq	[MFR-1]
External controller	approx 4 kg pack, two rechargeable batteries at about 5 h each	[TEAM-MAG-1]

The 60 by 70 mm figure is from the magnetic bearing paper and is given as approximate. It is the only published overall envelope found. See finding 1 above before using it with the volute dimensions in Part 1.

Right side of the rotor, all [PATENT-1][PATENT-2]
Parameter	Value	Claimed range
Outer vane diameter	approx 24 to 25 mm	22 to 27 mm
Primary vane inner diameter	16 mm	14 to 18 mm
Secondary vane inner diameter	approx 19 mm	18 to 20 mm
Vane height	approx 17.5 mm	15 to 20 mm
Vane thickness	1.5 mm	0.75 to 2.5 mm
Vane edge fillet	0.25 to 1.14 mm
Vane inlet and outlet angles	approx 90 deg both	80 to 100 deg
Outlet throat area	approx 150, 175 or 200 mm sq quoted	130 to 250 mm sq
Outlet width	8 to 12 mm
Cutwater angle	approx 70 deg	45 to 90 deg
Vane passage minimum flow area	min 650 mm sq	500 to 1500 mm sq
Pressure rise at 6 L/min	approx 20 mmHg	10 to 30 mmHg
Axial sensitivity	200 um axial shift changes flow by 0.5 to 1.5 L/min, or head by at least 10 mmHg

Rotor, cavity, gaps
Parameter	Value	Source
Rotor height	approx 10 mm, 6 to 13 mm claimed	[PATENT-1][PATENT-2]
Inlet port diameter	approx 18 to 20 mm, 18 to 22 mm claimed	[PATENT-1][PATENT-2]
Central cavity diameter	approx 28 mm, 27 to 29 mm claimed	[PATENT-1][PATENT-2]
Rotor to cavity wall radial gap	approx 4 mm average	[PATENT-1][PATENT-2]
Axial gaps	250 to 350 um, typically 300 um	[PATENT-2]
Clearance gaps in operation	at least 240 um during normal operation, falling to 100 um under extremes	[PAPER-4]
Clearance gaps, engineers' statement	at least 240 um in normal operation, over 20 times a red blood cell	[TEAM-MAG-1]
Motor air gap	3.4 mm, described as very wide	[PAPER-1]
Shunt flow path area	approx 25 mm sq, 15 to 50 mm sq claimed	[PATENT-1][PATENT-2]
Minimum flow path cross section	at least 50 mm sq	[PATENT-1]
Volute maximum radial force	under 0.85 to 1.2 N	[PATENT-1][PATENT-2]

The 240 um working clearance is the most transferable number in this set. It is the figure the BiVACOR team give as their hemocompatibility margin, and two independent sources agree on it. See finding 3 above for the tension with the patent axial gap.

Suspension and motor

All [PAPER-1] unless marked.

Three axial magnetic bearings, evenly spaced 120 deg apart, each a horseshoe shaped electromagnet in the casing with a coil on one leg.
Two ring permanent magnets on the rotor disk, inner and outer, producing bias flux across the air gap.
Three eddy current displacement sensors, arranged between the bearings, measuring axial displacement.
Axial flux brushless DC motor: twelve stator poles, eight permanent magnets on the rotor, run as a permanent magnet synchronous motor using back EMF for rotor position.
The left impeller sits in the motor side air gap, which is why that gap is 3.4 mm.
Radial support is passive hydrodynamic, not active [PAPER-2].
Axial rotor travel: plus or minus 0.3 mm in position control mode [PAPER-1]. An earlier development bearing reported plus or minus 0.15 mm [PAPER-2]. These are different hardware four years apart, not a contradiction.
Suspension power consumption under 15.5 W in the earlier development bearing [PAPER-2]. That is a 2010 development bearing, not the implanted device.
Virtual zero power control holds average magnetic bearing current near a 0.1 A set point [PAPER-1].

Operating envelope
Quantity	Value	Source
Speed, stable levitated range in animal trial	1800 to 2800 rpm	[PAPER-1]
Speed, as reported by the engineering team	1600 to 2700 rpm	[TEAM-MAG-1]
Speed, continuous flow bench test point	2400 plus or minus 100 rpm	[PAPER-4]
Speed, patent implantation protocol	1000 to 1250 rpm at start, then approx 1800 rpm	[PATENT-1][PATENT-2]
Design flow range	2 to 8 L/min, up to 12 L/min in exercise	[PATENT-1]
Flow claimed per side	over 12 L/min, marketing register, treat as a ceiling claim	[TEAM-MAG-1][MFR-1]
Bench hemocompatibility point	5.0 plus or minus 0.2 L/min, left 100 plus or minus 4 mmHg, right 27 plus or minus 4 mmHg	[PAPER-4]
Off design bench points	3.0 L/min low, 12.0 L/min high	[PAPER-5]
Pulsatile mode	speed modulation down to 1200 rpm, giving 40 to 80 mmHg pulse pressure	[PAPER-4]
Pulsatile mode, alternative	plus 900 rpm at 1 Hz	[PAPER-5]
Washout pulse	700 rpm once per minute	[PAPER-4]
Blood volume in the device loop	0.42 L total for both sides, approx 0.21 L per side. This is loop volume, not device priming volume.	[PAPER-4]
Flow sensitivity	10 mmHg pressure change gives at least 2 L/min flow change	[PATENT-1][PATENT-2]
Hemolysis, normalised index	0.004 plus or minus 0.003 g/100 L continuous and pulsatile, against 0.003 plus or minus 0.002 for the clinical reference pump	[PAPER-4]
Early in vitro result, 2008 concept	restored flow from 2 L/min pathological to 5 L/min, maximum left to right outflow differential 1.8 L/min	[PAPER-3]

Anatomical fit [PAPER-6]

Multicentre fitting study, ten transplant patients, using a 1:1 titanium model with correct port configuration.

Mean spine to sternum distance at T10: 14 cm, range 11 to 18 cm.
Mean distance from aorta to the aortic port: 0.2 cm, range 0 to 0.5 cm.
Mean distance from pulmonary artery to its port: 4.2 cm, range 1 to 7 cm.
The device fitted in all ten chests, across a range of body morphologies and heart failure aetiologies.

This study exists as a conference abstract and as a full paper, reporting slightly different measurements. The journal version is used here and the two are not mixed.

Real world total artificial hearts, for comparison

SynCardia CardioWest TAH-t, from the FDA approval document [REG-1]:

Two artificial ventricles, semi rigid polyurethane housing, four flexible polyurethane diaphragms per ventricle separating blood chamber from air chamber.
Mechanical valves in the inflow, 27 mm, and outflow, 25 mm, ports of each ventricle.
Maximum dynamic stroke volume 70 mL per ventricle, giving up to 9.5 L/min.
Contraindicated below 1.7 m sq body surface area, or with a sternum to tenth anterior vertebral body distance under 10 cm on CT.
Nominal console settings: left drive pressure 180 to 220 mmHg, right drive pressure 50 to 70 mmHg, rate 110 to 130 bpm, percent systole 50 to 55, diastolic vacuum 8 to 12 mmHg, giving average output 6.5 to 7.5 L/min.
Working stroke volume is set to 50 to 65 mL on the monitoring computer, below the 70 mL maximum, so that the ventricle never fully fills.

CARMAT Aeson, from a 2025 review [PAPER-7]:

Maximum ventricle volume 60 mL, stroke volume 55 to 60 mL.
Flow 2 to 9 L/min, rate 35 to 150 bpm.
Four biological valves.
8 mm diameter transmission cable, external control system about 4 kg, external battery about four hours.

Contrast worth carrying into our own work: SynCardia and Aeson buy pulsatility with displacement volume and valves, and pay for it in size and in wear parts. BiVACOR buys size and part count with a rotary pump, and has to synthesise pulsatility from speed modulation.

Sources

References

Full citations, with the source keys used throughout this page. The complete list, together with the method and fact by fact provenance, is held in BIVACOR-REFERENCE.md and BIVACOR-RESEARCH-REPORT.md, held in ~/cardianova/source(nont)/.

[PAPER-1] Kurita N, Timms D, Greatrex N, Kleinheyer M, Masuzawa T. Optimization design of magnetically suspended system for the BiVACOR total artificial heart. ISMB14, 14th International Symposium on Magnetic Bearings, Linz, Austria, 11 to 14 August 2014, pp 437 to 440.

[PAPER-2] Greatrex NA, Timms DL, Kurita N, Palmer EW, Masuzawa T. Axial magnetic bearing development for the BiVACOR rotary BiVAD/TAH. IEEE Trans Biomed Eng. 2010;57(3):714 to 721. doi:10.1109/TBME.2009.2033389. PMID 19822465.

[PAPER-3] Timms D, Fraser J, Hayne M, Dunning J, McNeil K, Pearcy M. The BiVACOR rotary biventricular assist device: concept and in vitro investigation. Artif Organs. 2008;32(10):816 to 819. doi:10.1111/j.1525-1594.2008.00633.x.

[PATENT-1] US 10,543,301 B2 and [PATENT-2] US 11,833,341 B2. Inventor Daniel Timms, assignee BiVACOR Inc.

[REG-1] FDA P030011, SynCardia CardioWest TAH-t. [REG-2] FDA PMA P160054, HeartMate 3, cited only as a negative result: it contains no dimensions.

Remaining keys, [PAPER-4] through [PAPER-13], [TEAM-MAG-1] and [MFR-1], are listed in full in BIVACOR-REFERENCE.md section 7.

BiVACOR: sourced dimensions and operating data