Linear-Spring Dashpot Model Overview

• The linear-spring dashpot model is a mechanical framework combining Hookean springs and Newtonian dashpots to represent linear viscoelastic behavior.
• It models dynamic responses in applications such as oscillatory systems, respiratory motion, and granular contact mechanics using a calibrated spring-dashpot network.
• Its simplicity enables reduction of complex continuum theories into tractable models, though careful calibration is needed for nonlinear and advanced viscoelastic phenomena.

The linear-spring dashpot model is a class of linear mechanical and viscoelastic models built from Hookean springs and Newtonian dashpots connected in series and/or parallel, so that forces or stresses are linear in displacement or strain and in velocity or strain rate. In its single-degree-of-freedom form it yields the mass–spring–damper equation $m\ddot{x}+c\dot{x}+kx=F(t)$; in constitutive form it yields Kelvin–Voigt, Maxwell, Zener, and higher-order linear viscoelastic networks. The same formalism appears in contact mechanics, discrete element methods (DEM), tissue and polymer rheology, vehicle suspensions, impact dynamics, and clinically motivated motion models such as the Voigt description of respiratory-driven lung tumor motion (0911.0137, Holm, 2017, Ackerley et al., 2012).

1. Canonical elements and constitutive structure

At the element level, the model consists of a linear spring with force law

$F_s = k\,x$

and a linear dashpot with force law

$F_d = c\,\dot{x},$

where $k$ is stiffness, $c$ is damping coefficient, $x$ is relative displacement, and $\dot{x}$ is relative velocity. In vibration problems this produces the standard linear equation

$m\ddot{x}(t)+c\dot{x}(t)+k x(t)=F(t),$

while in viscoelastic constitutive form the Kelvin–Voigt model gives

$$\sigma(t)=E_e\,\epsilon(t)+\eta\,\frac{\partial \epsilon(t)}{\partial t}.$$

The Zener, or standard linear solid, augments this by placing a Maxwell element in parallel with a spring, leading to

$$\sigma(t)+a_1\frac{\partial \sigma(t)}{\partial t}=E_e\,\epsilon(t)+\eta\,\frac{\partial \epsilon(t)}{\partial t}.$$

A general linear viscoelastic medium may equivalently be written in hereditary, differential, or relaxation-spectrum form, with a physically realizable relaxation modulus arising from the Maxwell–Wiechert model, a parallel network of Maxwell branches plus possibly a pure spring and/or pure dashpot (Ozcan et al., 2023, 0911.0137, Holm, 2017).

This elementary taxonomy is central because different topologies encode distinct physical responses. A Kelvin–Voigt element is spring and dashpot in parallel; a Maxwell element is spring and dashpot in series; a Zener model is a Maxwell element in parallel with a spring. In the cited elastography work, these are treated not as merely pedagogical analogies but as the common linear backbone underlying higher-order viscoelastic, fractional, poroelastic, and poroviscoelastic descriptions (Holm, 2017).

2. Dynamic-system interpretation and response characteristics

When written as an ordinary differential equation, the linear spring–dashpot model is a linear time-invariant second-order system. In the lung tumor motion model based on a Voigt unit, the governing equation is

$F_s = k\,x$0

where $F_s = k\,x$1 is the modeled tumor displacement, $F_s = k\,x$2 is the measured abdominal surrogate signal, $F_s = k\,x$3 is the spring parameter, $F_s = k\,x$4 is the damping parameter, and $F_s = k\,x$5 is a scale factor. In state-space form,

$F_s = k\,x$6

Its transfer function is

$F_s = k\,x$7

so the model acts as a damped, driven oscillator with amplitude filtering and phase lag (Ackerley et al., 2012).

In that clinical setting, the model was explicitly reported to handle baseline drift, amplitude changes, frequency variations, and phase shifts between internal and external respiratory motion. The dashpot contributes smoothing and phase lag, while the spring term preserves tethering to the driving displacement. The same damped-oscillator interpretation appears in droplet impact on a spring-supported plate, where the plate obeys

$F_s = k\,x$8

or in nondimensional form

$F_s = k\,x$9

with natural frequency $F_d = c\,\dot{x},$0, damping ratio $F_d = c\,\dot{x},$1, and critical damping $F_d = c\,\dot{x},$2 (Negus et al., 2020).

A closely related contact-dynamics form is obtained for a ball on a rigid surface under gravity: $F_d = c\,\dot{x},$3 which shows that even the simplest linear contact oscillator changes character when a constant forcing term is present. In that case the coefficient of restitution becomes impact-velocity dependent once gravity is retained, rather than being constant as in gravity-free treatments (Bartz, 2022).

3. Viscoelastic networks, generalized equivalents, and coarse-grained descriptions

The linear spring–dashpot model is also a reduction framework. In fractional viscoelasticity, the constitutive law remains linear but uses fractional derivatives; the cited elastography analysis shows that such models can be written as superpositions of ordinary spring–dashpot elements weighted by long-tailed relaxation-time or frequency spectra. For the fractional Zener model, the creep time-spectral function

$F_d = c\,\dot{x},$4

and the corresponding frequency-spectral function $F_d = c\,\dot{x},$5 imply a continuum of Kelvin–Voigt or Maxwell elements rather than a single discrete relaxation time. The same paper shows that Biot poroelastic shear waves are exactly equivalent to a three-element Zener model, and that the poroviscoelastic BICSQS extension is equivalent to a non-standard four-parameter spring–dashpot network (Holm, 2017).

An analogous coarse-graining appears in epithelial mechanics. Under small oscillatory deformations, the vertex model for epithelial tissues is fitted by standard linear spring–dashpot networks: the shear response in the solid phase is described by a standard linear solid, while the fluid phase is described by a Burgers model composed of two Maxwell elements in parallel. For bulk deformation, the solid hexagonal phase is purely elastic, whereas the fluid phase and disordered tilings are well described by a standard linear solid. Near the solid–fluid transition, additional Maxwell branches are required, indicating a broader relaxation spectrum rather than failure of the spring–dashpot paradigm itself (Tong et al., 2021).

This broader equivalence has two implications. First, spring–dashpot models are not restricted to literal mechanical devices; they function as reduced-order representations of far more complicated continuum theories. Second, the apparent simplicity of a linear spring–dashpot law can conceal a nontrivial distribution of relaxation processes, especially in fractional or disordered systems. This suggests that the model’s utility often lies in its spectral interpretation as much as in its element-level mechanics.

4. Contact mechanics, DEM, and restitution

In DEM and impact mechanics, the linear-spring dashpot model is most commonly used as a contact law. For equal-sized spheres in a vibrofluidized granular system, the normal interaction is

$F_d = c\,\dot{x},$6

with overlap

$F_d = c\,\dot{x},$7

and tangential interaction is a Coulomb-limited linear spring,

$F_d = c\,\dot{x},$8

For this LSD contact, the damping–restitution relation is

$F_d = c\,\dot{x},$9

In the cited simulations, the choice of tangential-to-normal stiffness ratio $k$0 had macroscopic consequences: $k$1 and $k$2 produced qualitatively different energy partition between translational and rotational motion, and equipartition was not observed for all realistic parameter values (Tiwari et al., 30 Aug 2025).

A 2026 analytical treatment generalizes this picture by showing that the power-law damped contact oscillator

$k$3

is exactly equivalent in phase space to a linear spring–dashpot oscillator under the transformation

$k$4

The resulting effective damping ratio is

$k$5

and the coefficient of restitution is exactly independent of impact velocity for all $k$6. The universal calibration formula is

$k$7

For $k$8 this reduces to the classical linear spring–dashpot result, and for $k$9 to the Hertzian spring–dashpot calibration of Antypov and Elliott (Feng, 29 Mar 2026).

The apparent simplicity of linear contact laws becomes more delicate in three dimensions. A detailed 3D DEM algorithm for the linear-frictional contact model shows that within one time-step it is necessary to account for changing tangential-force direction, transition from elastic to slip behavior, possible sliding during only part of the time-step, and twirling and rotation of the tangential force. Without three of these adjustments, errors are introduced in the incremental stiffness of an assembly; without the fourth, the resulting stress tensor is incorrect and is no longer a tensor. The same work gives explicit formulas for reversible and irreversible work increments during a time-step (Kuhn et al., 2020).

Rolling resistance can likewise be cast in spring–dashpot form. A recent single-parameter rolling model defines the rolling resistance moment as

$c$0

and derives

$c$1

from a single physically meaningful parameter, the critical rolling angle $c$2. The same study gives the stability condition

$c$3

for DEM time integration (Widartiningsih et al., 6 Feb 2026).

5. Representative applications

A clinically important example is respiratory motion tracking in lung radiotherapy. There the superior–inferior tumor displacement $c$4 is driven by a scaled abdominal signal $c$5 through the Voigt equation given above. The parameter triplet $c$6 was optimized by minimizing the root mean square error

$c$7

Across 52 treatment beams from 10 patients, Stage 1 beam-specific optimization produced a mean RMS residual error of $c$8 mm, with mean correlation coefficient $c$9. In Stage 2, patient-specific parameters taken from a single beam and reused across other beams increased the average RMS error from $x$0 mm to $x$1 mm, a mean increase of $x$2 mm, while the mean correlation coefficient remained $x$3. The model output was within an absolute bound of $x$4 mm and $x$5 mm for 95% of the time in Stage 1 and Stage 2, respectively (Ackerley et al., 2012).

Vehicle dynamics provides another canonical use. In a quarter-car front suspension, the linear spring–dashpot laws

$x$6

lead to

$x$7

$x$8

The cited optimization study uses these linear spring–dashpot models as the baseline before introducing prescribed nonlinear spring and damper characteristics, and evaluates ride comfort with weighted RMS body accelerations and handling with tire-force variation and roll-angle RMS (Ozcan et al., 2023).

In fluid–structure interaction, droplet impact onto a spring-supported plate reduces the substrate to a linear mass–spring–dashpot oscillator driven by the hydrodynamic force of the droplet. The plate undergoes forced damped oscillations, and the cited analysis and DNS show strong influence of plate mass, dashpot resistance, and spring stiffness on both plate motion and the droplet’s internal pressure field (Negus et al., 2020).

Polymer dynamics supplies a stochastic nonequilibrium application. A one-dimensional free-draining Hookean spring–dashpot polymer model with spring constant $x$9, bead friction $\dot{x}$0, and dashpot coefficient $\dot{x}$1 yields a closed expression for the average dissipated work under constant-velocity pulling. In the extrapolated zero-solvent-viscosity limit and at high trap stiffness,

$\dot{x}$2

so the dashpot coefficient can be estimated from the slope of dissipated work versus pulling velocity. For driven bead-spring-dashpot chains with $\dot{x}$3, by contrast, the dissipation is no longer expressible by a closed-form relation to a single dashpot coefficient; at high trap stiffness the dissipation decreases with $\dot{x}$4 for chains with internal friction, and the dependence on the internal-friction parameter becomes nonlinear as $\dot{x}$5 increases (Kailasham et al., 2019, Kailasham, 8 Jul 2025).

6. Limits, nonlinear departures, and mathematical issues

The linear spring–dashpot model is not universally adequate. A Bingham dashpot provides a direct counterexample: the force is not a function of velocity, while the velocity is a function of force. In the mass–spring–Bingham system,

$\dot{x}$6

must be supplemented by an algebraic constitutive relation for $\dot{x}$7, producing a semi-explicit differential-algebraic system rather than an ordinary differential equation. In free vibration, such a system can come to rest with a nonzero spring stretch, unlike the classical viscous dashpot case (0911.0137).

A related limitation appears under step inputs for nonlinear spring–dashpot networks. Classical distribution theory is sufficient for linear ODEs with Heaviside and Dirac inputs, but it fails for nonlinear constitutive laws because products and nonlinear functions of distributions are not defined in the classical framework. The cited work uses Colombeau algebra to analyze step load and step deformation in nonlinear spring–dashpot systems and recovers the linear standard linear solid as the special case

$\dot{x}$8

for which

$\dot{x}$9

This demarcates the boundary between linear spring–dashpot analysis and nonlinear generalized-function methods (Průša et al., 2016).

Even within linear elasticity, discreteness can introduce nontrivial structure. A recent spring-block construction derived from P1 finite elements produces tensor-valued spring constants $m\ddot{x}(t)+c\dot{x}(t)+k x(t)=F(t),$0 between blocks in 2D and 3D. The paper proves symmetry of these spring constants for homogeneous elasticity tensors and gives a necessary and sufficient condition for positive-definiteness in the isotropic case, as well as a sufficient condition in terms of mesh regularity and Poisson ratio. It further states that positive-definiteness of the spring constant derived from the finite element method plays a vital role in fracture simulations of elastic bodies using the spring-block system (Ounissi et al., 12 Sep 2025).

Application-specific limitations remain important. In lung tumor tracking, the cited study explicitly notes assumptions of linearity, a one-dimensional superior–inferior approximation, dependence on surrogate quality, and lack of explicit lead-time prediction (Ackerley et al., 2012). In granular contact, the choice of $m\ddot{x}(t)+c\dot{x}(t)+k x(t)=F(t),$1 can alter macroscopic predictions such as energy non-equipartition (Tiwari et al., 30 Aug 2025). In rigid-surface impacts, collision termination by zero compression can generate unphysical attractive forces, whereas force-based and hybrid termination criteria give different low-velocity restitution behavior (Bartz, 2022). This suggests that the linear-spring dashpot model is best understood as a calibrated linear backbone: physically interpretable, analytically tractable, and broadly extensible, but always conditional on constitutive invertibility, geometry, loading regime, and the fidelity required by the application.