Model Fidelity Research Articles

Characterization of Carbon-Black-Based Nanocomposite Mixtures of Varying Dispersion for Improving Stochastic Model Fidelity

Carbon black nanocomposites are complex systems that show potential for engineering applications. Understanding the influence of preparation methods on the engineering properties of these materials is critical for widespread deployment. In this study, the fidelity of a stochastic fractal aggregate placement algorithm is explored. A high-speed spin-coater is deployed for the creation of nanocomposite thin films of varying dispersion characteristics, which are imaged via light microscopy. Statistical analysis is performed and compared to 2D image statistics of stochastically generated RVEs with comparable volumetric properties. Correlations between simulation variables and image statistics are examined. Future and current works are discussed.

2-D Quantum Ising Model Fidelity and Order Parameter

This paper explores an interesting relationship between the ground-state fidelity quantum phase transitions and bifurcations per lattice site, and proposes a universal order parameter for the quantum Ising model for a square lattice of infinite-size in two spatial dimensions as a prototype model with symmetry breaking order. The approach is based on computing ground-state wave functions using a tensor network algorithm utilizing a representation with infinite projected entangled-pair states. The results can be applied to any systems with symmetry breaking order, because in the conventional Landau-Ginzburg-Wilson paradigm a quantum system subjected to a phase transition exhibits spontaneous symmetry breaking quantified by a local order parameter. A reduced fidelity bifurcation between two different reduced density matrices is also explored.

Development of a hybrid model for fuel performance analysis of spherical fuel elements under normal conditions

The spherical fuel element (SFE) dispersed with many tri-isotropic fuel (TRISO) particles has been primarily employed in the high-temperature gas-cooled reactor. Owing to the complex structure of SFE, behaviors will be evolved for both TRISO particles and the fuel matrix with the increasing burnup. To evaluate the fuel performance of SFE, the homogenization method has been conventionally adopted for SFE scale modeling, and dispersed TRISO particles are analyzed subsequently with fixed temperature boundary conditions. However, owing to the varying properties of TRISO particles under reactor operation, the homogenization models dedicated to certain circumstances can hardly reflect the properties evolution of TRISO with high fidelity, which may lead to uncertainties in analysis results. To this end, a hybrid model, namely the conventional homogenization model informed by the lower scale model online, has been developed in this paper, which may increase modeling fidelity for both SFE and particles. Specifically, property models for the TRISO particle and graphite matrix have been implemented in COMSOL. Subsequently, based on the linking of COMSOL and MATLAB, the hybrid model for SFE analysis was established. Through the hybrid model, the temperature field and stress distribution of TRISO, the evolution of effective properties, and the thermal and mechanical state of the SFE were obtained and analyzed. The results show that within the scope of this paper, the central temperature of TRISO in the innermost region rises with the increasing burnup for the decreasing thermal conductivity of materials, and the peak value is about 1324 K, which is about 60 K higher than that in the outermost region. The difference in the maximum plenum pressure is more than 5 MPa among TRISO particles, which is mainly caused by the variations in CO production. Consequently, the hoop stress of SiC in the outermost region is slightly elevated. The distribution of effective thermal conductivity (ETC) for TRISO is nearly 4.3 W/(m·K) at the beginning of life (BOL), while it was reduced to 3.6 W/(m·K) at the end of life (EOL). There are marginal variations along the radial direction for the ETC of TRISO. Meanwhile, the ETC of SFE decreases from about 21 W/(m·K) for fresh fuel to 13 W/(m·K) for spent fuel. The predicted effective elastic modulus and coefficient of thermal expansion of SFE increase gradually versus time. The former is caused by the densification of pyrolytic carbon, and the latter is mainly determined by the high volume fraction of matrix and the changes of CTE for pyrolytic carbon. In addition, the comparison in the temperature field between the proposed model and the Differential Effective Material Theory (D-EMT) model may further demonstrate the feasibility and effectiveness of the hybrid model.

Feedforward Control System Development and Implementation for Longitudinal and Lateral Motion of a Four-Wheeled Robot

The work presented in this paper advances through a series of physics-based increasing fidelity models that are used to design robot controllers that respect the robot's capabilities. It also develops a reference simple controller that is applicable to a large subset of tracking conditions, most of which involve non-invasive or highly dynamic movements, defines path geometry in response to the control problem, and develops both a simple geometric control and a dynamic model predictive control approach. In this research, we propose the mathematical modeling of the longitudinal and lateral movements utilizing PID with a feed-forward controller for a nonlinear model with disturbance impact. A feed-forward controller is suggested in this work to get rid of the disturbance effect.

Optical modeling of a radioisotope atmospheric chamber

Volatile radioisotopes represent a substantial health risk when released into the environment. To better understand the environmental fate of radioisotopes, the authors constructed a cylindrical steel-walled chamber to simulate the atmospheric processing of volatile radioactive gases. Optical modeling was performed for the properties of simulated sunlight in the chamber to better characterize atmospheric reaction studies. Optical simulations were performed using two wall materials (steel and thin-film silica) and validated against experimental measurements. This optical analysis methodology can be used to improve the fidelity of atmospheric models by accounting for optical inhomogeneities enabling a firmer grasp of radioisotopes’ environmental fate.Graphical abstract

An efficient process for macro geometry optimization of helical gear pairs considering static transmission error

Loaded tooth contact analysis (LTCA) adopted to obtain the static transmission error for the gear optimization affects significantly the computation time and results. This paper is to study an efficient process of macro geometry optimization of helical gear pairs with LTCA. An analytical LTCA model was proposed and verified with the measured transmission error. To find an efficient LTCA method, the proposed model was divided into three types (based on whether the slice stiffness and extended contact were included) by adjusting the fidelity. When extended contact was included in LTCA, the increased computation time was very short, but the accuracy of the results improved significantly. Optimization processes were designed based on the analysis results, and it was most efficient when starting with the model considering only the extended contact and ending 5%−10% of the entire generation with the highest fidelity model. In the process, the computational effort was significantly reduced compared to when the optimization process was configured with only the highest fidelity model, and the level of optimization was almost equal.

Dynamic modelling and simulation of industrial scale multistage flash desalination process

Multistage Flash (MSF) desalination process is still a dominant process, especially in the Gulf region, to produce high quality freshwater. Although there has been energy price surge in recent years, MSF process will continue to operate in that region for some foreseeable future. The key challenge is how to make such processes still profitable. Understanding the dynamics of any processes under uncertainty and disturbances is very important to make a process operationally feasible and profitable. The main aim of this work is to understand the dynamics of industrial scale MSF process using high fidelity and reliable process model. For this purpose, a detailed dynamic model for the MSF process incorporating key and new features is developed and validated against the actual data of a large-scale seawater desalination plant. The model is then used to study the behaviour of large scale MSF processes for disturbances in steam temperature, feed temperature and the recycle brine flow rate. The simulation results show that the last stage requires a longer time to settle compared to the preceding stages. In addition, steam temperature shows insignificant influence on the performance ratio compared to the inlet seawater temperature and recycle brine flow rate. Furthermore, it is found that the productivity of plant can increase in the winter compared to that in the summer. However, this benefit comes at the expense of increased steam consumption in the winter, resulting in a low performance ratio.

Considering the implementation research-to-practice gap: An experimental evaluation of intervention-general methods for assessing and supporting intervention fidelity through coaching

Implementation support through coaching—such as with embedded fidelity assessment, performance feedback, modeling, and alliance building—has been empirically supported as a way to increase and sustain interventionists' fidelity levels. However, education research consistently shows that practitioners struggle to monitor and improve interventionists' fidelity using implementation support strategies. One explanation for this type of implementation research-to-practice gap is that evidence-based coaching strategies have significant limitations with respect to their usability, feasibility, and adaptability. This study is the first to experimentally evaluate an evidence-based set of adaptable materials and procedures designed to assess and support the intervention fidelity of school-based interventions. Using a randomized multiple-baseline-across-participants design, we examined the extent to which these materials and procedures would influence intervention adherence and quality of an evidence-based reading intervention. Across all nine interventionist participants, data revealed that the implementation strategies meaningfully improved intervention adherence and quality, and high levels of intervention fidelity maintained 1 month after removing the support procedures. Findings are discussed with respect to how these materials and procedures address a critical need within school-based research and practice as well as how they may help to inform and address the implementation research-to-practice gap in education.

A BIM–LCA Approach for the Whole Design Process of Green Buildings in the Chinese Context

The integrated description of the building geometry and the element attributes of the building information model (BIM) can reduce the effort needed to acquire data for life cycle assessment (LCA) and life cycle costing (LCC) at each design stage while supporting their potential for analyzing life cycle performances and feeding back to the design process. To support this, several methods and tools have been proposed that aim to obtain the life cycle performances of buildings following the level of model fidelity with the life cycle inventory (LCI) database at different scales. However, inconsistencies in decision-making caused by regional differences in LCA/LCC data sources, benchmarks, and building standards cannot be ignored. In this study, a scalable LCA/LCC method integrated with the BIM platform is proposed for the whole green building design process in the Chinese context, and it is implemented with a developed tool based on Revit. A national-/regional-specified database of building elements and materials is established. Referring to China’s carbon-neutral target and relevant standards for green buildings, the baseline values are deduced, and a reference building is defined accordingly to facilitate the evaluation and improvement of the design scheme. According to the Assessment Standard for Green Building (GB50378-2019) and the survey of architectural design practices in China, the key parameters at different design stages are defined. The method and tool are demonstrated using the case study of a school building, analyzing its life cycle carbon emissions and life cycle costs throughout the design process. The results show that the proposed method can facilitate the improvement of the scheme at different design stages and that it can cope with different data accuracies and different LODs in the building information model in the Chinese green building design process. Lastly, the uncertainties raised by the data quality and time-associated factors are discussed.

Integrating plant physiology into simulation of fire behavior and effects.

Wildfires are a global crisis, but current fire models fail to capture vegetation response to changing climate. With drought and elevated temperature increasing the importance of vegetation dynamics to fire behavior, and the advent of next generation models capable of capturing increasingly complex physical processes, we provide a renewed focus on representation of woody vegetation in fire models. Currently, the most advanced representations of fire behavior and biophysical fire effects are found in distinct classes of fine-scale models and do not capture variation in live fuel (i.e. living plant) properties. We demonstrate that plant water and carbon dynamics, which influence combustion and heat transfer into the plant and often dictate plant survival, provide the mechanistic linkage between fire behavior and effects. Our conceptual framework linking remotely sensed estimates of plant water and carbon to fine-scale models of fire behavior and effects could be a critical first step toward improving the fidelity of the coarse scale models that are now relied upon for global fire forecasting. This process-based approach will be essential to capturing the influence of physiological responses to drought and warming on live fuel conditions, strengthening the science needed to guide fire managers in an uncertain future.

The Reflective Fostering Programme fidelity rating scale: development and inter-rater reliability

PurposeThe purpose of this study is to describe the development of the 14-item reflective fostering fidelity rating (RFFR), an observational rating system to evaluate model fidelity of group facilitators in the Reflective Fostering Programme (RFP), a mentalisation-based psychoeducation programme to support foster carers. The authors assess usability, dimensionality, inter-rater reliability and discriminative ability of the RFFR.Design/methodology/approachEighty video clip extracts documenting 20 RFP sessions were independently rated by four raters using the RFFR. The dimensionality of the RFFR was assessed using principal components analysis. Inter-rater agreement was assessed using the intra-class correlation coefficient.FindingsThe proportion of missing ratings was low at 2.8%. A single principal component summarised over 90% of the variation in ratings for each rater. The inter-rater reliability of individual item ratings was poor-to-moderate, but a summary score had acceptable inter-rater reliability. The authors present evidence that the RFFR can distinguish RFP sessions that differ in treatment fidelity.Originality/valueTo the best of the authors’ knowledge, this is the first investigation and report of the RFFR’s validity in assessing the programme fidelity of the RFP. The paper concludes that the RFFR is an appropriate rating measure for treatment fidelity of the RFP and useful for the purposes of both quality control and supervision.

Using Machine Learning for Loss Prediction in a Hybrid Meanline Modeling Method to Deliver Improved Radial Turbine Performance Prediction

Abstract Low fidelity modeling approaches remain attractive due to an unrivaled ability to predict full turbine performance maps quickly compared to high-fidelity approaches such as computational fluid dynamics (CFD), especially in the preliminary design process. As improvements in performance on a component level approach a point of diminishing returns, the ability to efficiently optimize the complete charging system for a given duty is a topic attracting significant research interest. In the case of turbocharging applications, existing engine and powertrain simulations require turbine maps to calculate the turbine performance, which are usually obtained from experimental testing. Unfortunately, the need for extrapolation is unavoidable because of the limited range of testing data available, leading to inaccuracies especially at off-design conditions. To enable intensive modeling and optimization of complete vehicle powertrains for different drive cycles, the current piece of work seeks to combine the advantages of machine learning techniques and physical meanline modeling to facilitate faster, more accurate predictions of complete turbocharger maps. This paper presents a novel methodology for turbocharger turbine rotor and nozzle performance prediction based on hybrid modeling. The turbine rotor and nozzle were parameterized to conduct CFD simulations for a wide variety of turbine geometries, which were used to form a database to train an artificial neural network (ANN). The predicted losses provided by the ANN were then utilized in the meanline code, substituting for the conventional empirical loss models. As well as removing the need for empirical loss models, modifications were undertaken to the meanline approach to further enhance modeling accuracy. First, in order to accurately characterize the stage mass flow capacity, the losses occurring in the nozzle and rotor were subdivided into those occurring before and after the throat. A second novel aspect is that the aerodynamic blockage level at the rotor throat was implemented as a variable rather than a constant value. By training the ANN to predict the variation of blockage with geometry and operating condition, a more accurate depiction of the changing secondary flow fields could be achieved. The capabilities of the hybrid meanline modeling method were evaluated on several unseen test cases. The resulting predictions of efficiency and mass flowrate demonstrated strong correlation with CFD results and experimental test results. The hybrid meanline modeling method therefore displays great potential in wide range radial turbine performance prediction with enhanced accuracy in comparison to traditional approaches.

A latent survival model integrated computer simulation-based evaluation for nursing home staffing

Nursing homes (NHs) are critical facilities for caring frail older adults with around-the-clock formal care and personal assistance. To ensure quality of care for NH residents, an adequate staffing level is of great importance. Current NH staffing practice is mainly based on experience and regulation. The objective of this paper is to investigate the viability of experience-based and regulation-based strategies, as well as alternative staffing strategies to meet the heterogeneous service demand of NH residents at reduced labor cost under various scenarios of census compositions. We propose a predictive analytics integrated computer simulation model to characterize the heterogeneous service demand of NH residents, and further evaluate and identify promising staffing strategies at the facility level. Specifically, we propose a predictive model based on latent survival analysis to characterize diverse length-of-stay (LOS) with multiple discharge dispositions among NH residents. Further, we develop a simulation model with the incorporation of predictive analytics and domain knowledge to characterize the heterogeneous service demand of NH residents on different types of caregivers over time. Based on the simulation model, we develop a graphical user interface for the simulator to evaluate different staffing strategies at the facility level and inform NH administrators about promising strategies. We use real NH data to validate the proposed model and demonstrate its effectiveness. The proposed predictive LOS model considering multiple discharge dispositions exhibits superior prediction performance and offers better staffing decisions at reduced costs than those without the consideration. With the improved modeling fidelity via integrating predictive analytics with computer simulation, the proposed model is flexible to evaluate various staffing strategies using total labor cost as a performance metric, and can identify promising staffing strategies to meet the service demand of NH residents. Promising staffing strategies with the suggested staff-to-resident (SR) ratio can significantly reduce the total labor cost of multiple types of caregivers, as compared to the benchmark strategies, such as the SR ratios based on industrial practice or minimum requirement of state regulation. Moreover, we construct multiple scenarios of different census compositions of NH residents to demonstrate the capability of the proposed model. Our proposed model can facilitate NH staffing decision making to meet the heterogeneous service demand of NH residents at reduced labor costs.

Identifiability of soft tissue constitutive parameters from in-vivo macro-indentation

Reliable identification of soft tissue material parameters is frequently required in a variety of applications, particularly for biomechanical simulations using finite element analysis (FEA). However, determining representative constitutive laws and material parameters is challenging and often comprises a bottleneck that hinders the successful implementation of FEA. Soft tissues exhibit a nonlinear response and are commonly modeled using hyperelastic constitutive laws. In-vivo material parameter identification, for which standard mechanical tests (e.g., uniaxial tension and compression) are inapplicable, is commonly achieved using finite macro-indentation test. Due to the lack of analytical solutions, the parameters are commonly identified using inverse FEA (iFEA), in which simulated results and experimental data are iteratively compared. However, determining what data must be collected to accurately identify a unique parameter set remains unclear. This work investigates the sensitivities of two types of measurements: indentation force-depth data (e.g., measured using an instrumented indenter) and full-field surface displacements (e.g., using digital image correlation). To eliminate model fidelity and measurement-related errors, we employed an axisymmetric indentation FE model to produce synthetic data for four 2-parameter hyperelastic constitutive laws: compressible Neo-Hookean, and nearly incompressible Mooney-Rivlin, Ogden, and Ogden-Moerman models. For each constitutive law, we computed the objective functions representing the discrepancies in the reaction force, the surface displacement, and their combination, and visualized them for hundreds of parameter sets, spanning a representative range as found in the literature for the bulk soft tissue complex in human lower limbs. Moreover, we quantified three identifiability metrics, which provided insights into the uniqueness (or lack thereof) and the sensitivities. This approach provides a clear and systematic evaluation of the parameter identifiability, which is independent of the selection of the optimization algorithm and initial guesses required in iFEA. Our analysis indicated that the indenter’s force-depth data, despite being commonly used for parameter identification, was insufficient for reliably and accurately identifying both parameters for all the investigated material models and that the surface displacement data improved the parameter identifiability in all cases, although the Mooney-Rivlin parameters remained poorly identifiable. Informed by the results, we then discuss several identification strategies for each constitutive model. Finally, we openly provide the codes used in this study, to allow others to further investigate the indentation problem according to their specifications (e.g., by modifying the geometries, dimensions, mesh, material models, boundary conditions, contact parameters, or objective functions).

In search of a data-driven symbolic multi-fluid ten-moment model closure

The inclusion of kinetic effects into fluid models has been a long standing problem in magnetic reconnection and plasma physics. Generally, the pressure tensor is reduced to a scalar which is an approximation used to aid in the modelling of large scale global systems such as the Earth's magnetosphere. This unfortunately omits important kinetic physics which have been shown to play a crucial role in collisionless regimes. The multi-fluid ten-moment model, however, retains the full symmetric pressure tensor. The ten-moment model is constructed by taking moments of the Vlasov equation up to second order, and includes the scalar density, the vector bulk-flow and the symmetric pressure tensor for a total of ten separate components. Use of the multi-fluid ten-moment model requires a closure which truncates the cascading system of equations. Here we look to leverage data-driven methodologies to seek a closure which may improve the physical fidelity of the ten-moment multi-fluid model in collisionless regimes. Specifically, we use the sparse identification of nonlinear dynamics (SINDy) method for symbolic equation discovery to seek the truncating closure from fully kinetic particle-in-cell simulation data, which inherently retains the relevant kinetic physics. We verify our method by reproducing the ten-moment model from the particle-in-cell (PIC) data and use the method to generate a closure truncating the ten-moment model which is analysed through the nonlinear phase of reconnection.

Modulation of potassium channels preserves temporal fidelity in auditory network models.

Potassium channels in auditory neurons are rapidly modified by changes in the auditory environment. In response to elevated auditory stimulation, short-term mechanisms such as protein phosphorylation and long-term mechanisms linked to channel synthesis increase the activity of channels that promote high frequency firing. We have now used simple simulations of cochlear hair cells and postsynaptic neurons to demonstrate that the amplitudes of potassium currents in neurons required for optimal encoding of a low-level auditory signal differs substantially from that for louder sounds. Specifically, the cross correlation of the output of a neuron with an auditory stimulus is increased by increasing potassium currents as sound amplitude increases. Although it has been suggested that channel modulation allows neurons to fire at high rates in response to high sound levels, we found that the cross correlation is entirely independent of firing rate and that combinations of currents that maximize firing to a stimulus provide very poor temporal fidelity. We also found that levels of potassium currents that maximize the temporal fidelity of the output of an ensemble of thirty neurons differ from those that maximize temporal fidelity for a single neuron. This suggests that, to maximize preservation of temporal information, modulatory mechanisms must coordinate channel activity in groups of neurons receiving similar synaptic inputs. The simulations provide an explanation for the modulation of the intrinsic excitability of auditory brainstem neurons by changes in environmental sound levels, and the results may extend to information processing in other neural systems.

Automation of a National Noise Model

The United Kingdom Department for Food and Rural Affairs (Defra) commissioned the design and build of an environmental noise modelling system (NMS). The NMS has been developed to support Defra in developing its environmental noise evidence base by preparing national road and railway noise models. This paper discusses the automation aspects of the NMS with respect to the various elements required for a noise model. This includes the development of sophisticated data processing engines and the techniques utilised to automate the delivery of a 3-dimensional noise propagation environment, and road and railway noise emissions. It highlights several innovative methods which are now possible due to the data landscape in England and how through open data standards the NMS has the potential to improve the fidelity of national noise models.

Improved-Efficacy EM-Driven Optimization of Antenna Structures Using Adaptive Design Specifications and Variable-Resolution Models

Optimization-driven parameter tuning is an essential step in the design of antenna systems. Although in many cases, it is still conducted through parametric studies, rigorous numerical methods become a necessity if truly optimum designs are sought for, and the problem intricacies (number of variables, multiple goals, and constraints) make the interactive approaches insufficient. The two practical considerations of electromagnetic (EM)-driven optimization are reliability and computational cost. Repetitive EM simulations may incur unmanageable expenses, whereas the lack of a decent starting point or objective function multimodality may prevent the numerical procedures (especially the local ones) from identifying satisfactory designs. In pursuit of reliability improvements, a design specification adjustment procedure has been recently proposed that improves the immunity of local search procedures to poor starting points, where the objective function is modified by relocating the design goals (e.g., center frequencies) closer to the actual operating parameters of the antenna at the current design, to make them attainable through local search. The goals are then gradually adjusted and converged to the original targets toward the end of the optimization process. In this article, we introduce an algorithm for reliable and low-cost antenna tuning that capitalizes on the specification management scheme while embedding it in variable-resolution optimization framework. In our approach, the EM model fidelity is adaptively adjusted based on the misalignment between the actual and target operating conditions, as well as the convergence status of the algorithm. By initiating the search process from the lowest fidelity model (gradually dialed up to the highest fidelity one), considerable computational savings of almost 60% can be achieved, with respect to the single-fidelity procedure. The speedup is possible without compromising reliability of the optimization process, as demonstrated using three examples of microstrip antennas, designed under different and challenging scenarios.

3-D EM Modeling of Medical Microwave Imaging Scenarios With Controllable Accuracy

Medical microwave imaging (MMWI) is a multidisciplinary task that requires precise electromagnetic modeling of heterogeneous human tissues, solving non-linear and ill-posed inverse scattering problems with weak contrasts, and design of compact antenna arrays well-matched to adjacent body parts. If any of the activities listed above does not perform well, the resulting imaging will be poor. In this article, we devote a special effort to achieving high accuracy in electromagnetic simulations of MMWI scenarios. This complex task requires developing a suitable meshing procedure, which balances model fidelity and computational bottlenecks such as a large number of unknowns in the electromagnetic model. Thus, instead of a simple conversion from triangles to quads, we offer a technique that generates high-quality mesh (in terms of both simulation speed and geometrical accuracy) with controllable deviation from the initial model. In addition, we developed a strategy for self-convergence tests that prove the reliability of the simulation results in the MMWI framework. Moreover, an additional validity test has been proposed, based on the inverse scattering operator, commonly used in many microwave imaging algorithms. The effectiveness of these novel techniques is demonstrated with the example of a 21-antenna system placed around the 5-tissue head phantom with and without stroke.

Optimizing autoinjector devices using physics-based simulations and Gaussian processes

Autoinjectors are becoming a primary drug delivery option to the subcutaneous space. These devices need to work robustly and autonomously to maximize drug bio-availability. However, current designs ignore the coupling between autoinjector dynamics and tissue biomechanics. Here we present a Bayesian framework for optimization of autoinjector devices that can account for the coupled autoinjector-tissue biomechanics and uncertainty in tissue mechanical behavior. The framework relies on replacing the high fidelity model of tissue insertion with a Gaussian process (GP). The GP model is accurate yet computationally affordable, enabling a thorough sensitivity analysis that identified tissue properties, which are not part of the autoinjector design space, as important variables for the injection process. Higher fracture toughness decreases the crack depth, while tissue shear modulus has the opposite effect. The sensitivity analysis also shows that drug viscosity and spring force, which are part of the design space, affect the location and timing of drug delivery. Low viscosity could lead to premature delivery, but can be prevented with smaller spring forces, while higher viscosity could prevent premature delivery while demanding larger spring forces and increasing the time of injection. Increasing the spring force guarantees penetration to the desired depth, but it can result in undesirably high accelerations. The Bayesian optimization framework tackles the challenge of designing devices with performance metrics coupled to uncertain tissue properties. This work is important for the design of other medical devices for which optimization in the presence of material behavior uncertainty is needed.