3D printed aortic valve phantom mechanics: what biaxial tensile testing shows before you ever run a flow loop

A 3D printed aortic valve phantom is easy to judge by eye. It is harder to tell (before you run a flow loop) whether the wall and leaflets are already operating in a stiff regime. In a study out of the University of Toronto and Toronto Metropolitan University, the authors started with equi-biaxial tensile testing on a CellScale BioTester, tracking strain optically in the center of each specimen, and only then moved on to pressure-drop measurements in printed adult and pediatric models. The workflow is simple on paper, but it catches a lot of the “tissue-mimicking” assumptions people usually leave untested.

The team compared Stratasys TissueMatrix materials and silicone elastomers against native human ascending aorta. Then they printed an idealized aortic geometry with an integrated semilunar valve and looked at how pressure drop scaled with flow.

If you work on cardiovascular phantoms, this reads less like a single result and more like a practical warning: stiffness choices that seem minor at the printer can show up again when you put the model under load. Let’s jump into the research.

3D printed aortic valve phantom validation usually breaks at the material assumption

When a 3D printed aortic valve phantom gets described as “tissue-mimicking,” it is often shorthand for something much looser (i.e., “it is flexible” or “it prints without tearing”). The paper pushes on a more concrete version of that claim. If the goal is to use a phantom for hemodynamic testing or surgical training, it matters whether the printed wall and valve are behaving like human aortic tissue biomechanics, not just whether the geometry looks anatomical.

They focused on the ascending aorta and kept the tissue baseline realistic rather than tidy. Tissue came from 20 donors, with samples taken from the inner and outer curvature regions, and tested along both circumferential and longitudinal directions. That matters because the ascending aorta is direction-dependent, and even within a single vessel you can measure stiffness differences that are not small.

Instead of forcing the aorta into a single “elastic modulus,” they used tangent moduli as a way to describe the nonlinear response. They extracted a low tangent modulus in the early part of the curve, and a higher tangent modulus at larger strains. It is not a constitutive model, but it is enough to show where a material is operating on the stress-strain curve.

This is where the study starts to feel useful for phantom work. If your printed material looks linear and your native tissue stiffens as strain increases, you can already see the mismatch coming.

We wrote about another valve-focused workflow recently, where small design choices show up quickly in bench testing.

Biaxial tensile testing on a BioTester: how they measured strain, not just load

A lot of phantom materials get compared with uniaxial coupons, sometimes even with a single pull direction. That can be convenient, but it is not how an aortic wall sees load. Here, the authors used biaxial tensile testing to create a closer match to physiological loading, then used the same approach for tissue, TissueMatrix, and silicone.

They used a CellScale BioTester in an equi-biaxial configuration with tungsten rakes (our patented BioRakes™️). A small but important detail is how they handled strain measurement. Instead of relying on actuator displacement, they placed four tracking dots near the center of the specimen and used the BioTester’s integrated camera to track deformation in that central region.

If you have ever watched a rake-mounted sample under load, you know why that matters. Grips and attachments can dominate the deformation field, especially for compliant materials. The center-of-sample tracking keeps the comparison focused on what the material itself is doing, not what the attachment points are doing to it.

If you’re newer to soft material testing and want a plain-language primer before getting deep into biaxial data, this overview is helpful.

A fixture problem that becomes a material problem

One thing that stands out in the paper is that the printed materials did not only behave differently because of their modulus. Some TissueMatrix samples tore or ruptured at the rake attachment sites when pushed to higher strains, which disrupted strain tracking. Silicone elastomers did not show the same tearing behaviour even at higher strains.

It is an annoyingly practical point, but it is also relevant to anyone trying to build a 3D printed aortic valve phantom workflow. You can have a material that feels compliant, but if it cannot survive your test fixture, you may never get meaningful biaxial data out of it. That can push labs toward lower strain comparisons and linear fits, even when the tissue they are trying to match is nonlinear.

Stratasys TissueMatrix mechanical characterization versus human aortic tissue biomechanics

For the printed materials, the authors focused on Stratasys TissueMatrix options intended for soft-tissue applications. Their Stratasys TissueMatrix mechanical characterization was done under equi-biaxial loading, but limited to 5% engineering strain for synthetics. That limit is partly methodological (the printed materials behaved linearly in that region), and partly pragmatic (attachment tearing becomes a factor as strain increases).

The main comparison is straightforward: TissueMatrix samples were significantly stiffer than native ascending aorta when compared to the tissue’s tangent modulus values. Even without quoting exact numbers, the direction of the mismatch is clear. The “tissue-mimicking” print materials are operating in a stiffer regime than a healthy aorta, at least under biaxial loading in the tested range.

They also report anisotropy within the print plane, with stiffness differences between X and Y directions for TissueMatrix. That is easy to overlook in a phantom build, because rotating a model on the build plate feels like a geometry decision, not a mechanics decision. But in practice it can change the directional response of a printed wall or leaflet.

Silicone elastomer vascular models as a reality check

The silicone part of the study is a useful counterweight. Silicone is often treated as the fallback option when printing does not give the compliance you want. In these tests, silicone elastomer vascular models were not significantly different from native aorta in modulus comparisons, while TissueMatrix was.

That does not mean silicone automatically solves phantom realism. Silicone parts are harder to integrate into complex, anatomically faithful geometries unless you mold them or use hybrid workflows. But it does complicate the default assumption that 3D printing is inherently closer to tissue because it is “advanced.”

If you are building a 3D printed aortic valve phantom and seeing a stiff functional response, silicone is a reminder that mechanical compliance can come from simpler materials, even if the fabrication workflow becomes more manual.

Flow loop testing of a 3D printed aortic valve phantom: pressure drop versus Reynolds number

After mechanical testing, the authors printed adult and pediatric aortic phantoms with an integrated semilunar valve and tested them in a mock circulation loop. Pressure was measured proximal and distal to the valve, and pressure drop was computed across the valve over a range of flows, reported in terms of Reynolds number.

This is where the study shifts from “what does the material do under controlled biaxial loading” to “what does the printed valve do when you actually run fluid through it.”

If you want a separate read on what makes valve tissue tricky to match mechanically, this post on leaflet mechanics is a good place to start

What you see is a clean scaling: pressure drop increases with Reynolds number for both adult and pediatric phantoms. The valves remain intact over the tested conditions. But the more interesting detail is that the “healthy” and “stiff” material cases track fairly closely. You do not get a dramatic separation just by toggling between compliant and stiffer TissueMatrix options.

There are a few ways to interpret that. It may be that the stiffness difference between the two printed sets is not the dominant driver of valve opening in this geometry. It may be that leaflet geometry and print fidelity matter more than the modulus shift the authors introduced. Or it may be that both TissueMatrix choices are already stiff relative to human tissue, so the tuning range is narrower than it appears.

The paper does not fully settle that question, but it leaves you with a reasonable caution: if your 3D printed aortic valve phantom behaves like a stiff valve, it may not be enough to just pick the softer material preset and assume you have moved into “healthy.”

BioTester use in the study: what it contributed beyond a number

The BioTester is not a side detail in this paper. It is what allowed them to put tissue, TissueMatrix, and silicone into the same mechanical frame without leaning on vendor data sheets or uniaxial proxies.

In this study, the BioTester was used to:

• apply equi-biaxial loading to soft tissue and soft polymers using rake attachments

• measure forces while tracking central deformation with simple optical markers

• compare linear-elastic synthetic behavior to nonlinear tissue behavior using consistent strain measurement

• expose practical fixture issues, like attachment tearing, that can shape how far you can push printed elastomers under biaxial tension

That combination is what makes the mechanical results feel actionable for phantom development. The point is not that one material is “good” and another is “bad.” It is that you can measure where your candidate materials sit relative to tissue before you commit to a full phantom print and a full hemodynamic test series.

BioTester as an instrument for 3D printed aortic valve phantom development

Most phantom development cycles are iterative and slightly messy. You print a geometry, it looks fine, you test it, it behaves stiff, you tweak the material assignment, and you try again. The frustrating part is that it can be difficult to tell whether you are fixing the right thing. Biaxial testing gives you a way to separate “material regime” problems from “geometry” problems early.

A biaxial system like the CellScale BioTester is typically used in this context to load soft specimens in two orthogonal directions while measuring force and tracking deformation optically. When you are dealing with compliant materials, central strain tracking is often the difference between data you can compare and data that mostly reflects boundary conditions.

For 3D printed elastomers, that matters in a very practical way. Printed materials can tear at attachments, show direction-dependent behavior, and behave linearly in one range and unpredictably in another. Those effects tend to show up in biaxial tensile testing before they show up as a dramatic failure in a flow loop. If you are trying to build a 3D printed aortic valve phantom that behaves like tissue, that early warning is useful.

Find out more about the BioTester and it’s capabilities here.

Read the Full Publication