Continuous Flow Vessels Research Articles

Optical diagnostics of the venting spray and combustion behaviour during Li-ion battery thermal runaway induced by ramp heating

• Thermal Runaway assessed by means of Natural Luminosity and Schlieren. • LCO, NMC and LFP cathodes are compared at different states of charge. • Similar venting penetration at different state of charge. • Combustion is observed in LCO and NMC cathodes. • LFP chemistry represents the safest operation among chemistries. Lithium-ion cells are one of the solutions with the most significant potential for energy storage in future transportation. However, a controlled ambient temperature is crucial for its application due to thermal safety. Thermal runaway characterisation of three types of commercially available cylindrical cells, based on cathode composition of LiCoO 2 (LCO), LiNi 0.6 Mn 0.2 Co 0.2 O 2 (NMC 622), and LiFePO 4 (LFP), is performed in the current investigation. The cells were preconditioned to achieve a state of charge (SOC) of 25%, 50%, and 100%; this allows for studying all the SOC operating windows used in electric vehicles. A continuous flow vessel is used as a testbed to monitor the cells going through ramp heating with clear visualisation. Schlieren and Natural Luminosity technique are applied with high-speed cameras. The cell surface temperature and the chamber temperature are measured with thermocouples. The chamber is filled with air that works as a heat transfer medium and chamber cleaner. The main novel of this work is high-quality images during both safety venting and combustion, which characterise the thermal runaway process in detail. High-speed cameras allow recording with more than 6,000 fps in both phenomena. The results show that the average temperature where the cells start the safety venting was found at 193°C. The exothermic onset was, on average, 238°C. The maximum temperature achieved in the cell surface was average for the nine experiments, 567°C. The high-speed camera recording shows that the LCO initiated the flame by a hot spot in the vent cap while the NMC ejected hot solid particles that burned the flammable gas. The LFP did not show any flame propagation for the three SOC tested.

Influence of environmental conditions in the battery thermal runaway process of different chemistries: Thermodynamic and optical assessment

Thermal runaway is one of the main concerns of battery electric vehicles due to the hazard level that represents for the user and the surroundings. Several works studied different type of abuse in lithium-ion cells and packs, but the understanding is still insufficient in terms of the combustion process. In this study, three different lithium-ion cell chemistries (LCO, NMC and LFP) are studied in two environmental conditions with different oxygen content (0 and 21%) in a continuous flow vessel to understand if the use of inert atmosphere may be a pathway to avoid thermal runaway. In addition, detailed optical research is conducted together with temperature sensing to understand the venting through the vent cap before the thermal runaway. The combustion is recorded with a high-speed camera (6000 fps) while the venting is visualized through a Schlieren technique with another high-speed camera (12,000 fps). The thermodynamic results show that the venting process can be detected by a cell surface temperature decrease of around 5 °C, while the thermal runaway is seen as a battery self-heating (cell temperature higher than the ambient) and a suddenly increase of temperature until 700 °C in the surface of the cell. The optical access to the combustion chamber allows to observe with detail the venting of liquid electrolyte amongst the gases generated by the thermal abuse. In addition, the combustion records show that with inert atmosphere the combustion it is not initiated, and the process is restricted to smoke ejection. By contrast, the case of air (21% O2) resulted in combustion outside the battery cell with high increase of the air temperature. In terms of battery chemistry, the Lithium, Ferrum, Phosphate (LFP) shows the highest safety time and lowest chamber temperatures. LCO and Nickel Manganese Cobalt (NMC) had similar behaviour in terms of safety time and temperature behaviour, but Lithium Cobalt Oxygen (LCO) shows more variation with respect to the atmosphere (reactive and inert) than NMC.

An optical investigation of thermal runaway phenomenon under thermal abuse conditions

This work aims at studying the thermal runaway process caused by thermal abuse using different optical techniques. A commercial Samsung ICR 18,650 – 26 J cylindrical battery cell is exposed to different heating ramps in a Continuous Flow Vessel (CFV) to identify critical phases of the battery thermal runaway. The open volume test bed allows to increase the temperature under controlled conditions and evacuate the released gases, enhancing the visualization of the battery venting and combustion. The venting of the electrolyte and gases is seen around 200 °C. The safety time, defined as time between venting and fire, was around 3 min, with an inverse relation with respect to the heating ramp. The use of fast cameras (schlieren, Natural Luminosity and OH*) allowed to deeply understand the phenomena of liquid electrolyte and generated internal gases venting as well as the combustion process.

Reduced uterine perfusion pressure decreases functional capillary density in skeletal muscle.

The purpose of this study was to examine the functional and structural capillary density in the reduced uterine perfusion pressure (RUPP) model, which when performed during pregnancy is an established animal model of preeclampsia. We hypothesized that the RUPP model would be associated with capillary rarefaction and impaired capillary perfusion, which would be more pronounced in the pregnant state. Female Sprague-Dawley rats (n = 32) were randomized to nonpregnancy (Nonpregnant) or breeding (Pregnant) at 12 wk of age and again to RUPP or SHAM surgeries on gestational day (GD) 14 (or equivalent age in nonpregnant rats). On GD 20 (or equivalent), capillary structure and perfusion of the extensor digitorum longus were imaged using digital intravital video microscopy. Functional videos were analyzed by a blinded observer to measure capillary density, expressed as capillaries per millimeter intersecting three staggered reference lines (200 μm). Flow was scored as the percentage of capillaries having 1) continuous, 2) intermittent, or 3) stopped flow. Total capillary density was not different between groups. There was a main effect of RUPP surgery resulting in decreased continuous flow vessels (P < 0.01) and increased stopped flow (P < 0.01), which was driven by differences between pregnant animals (Continuous flow: pregnant SHAM 80.1 ± 7.8% vs. pregnant RUPP 67.8 ± 11.2%, P < 0.05) (Stopped flow: pregnant SHAM 8.7 ± 3.2% vs. pregnant RUPP 17.9 ± 5.7%, P < 0.01). Our results demonstrate that the RUPP surgery is associated with a decrease in functional capillary density in skeletal muscle that is more pronounced in the pregnant state, which may contribute to the vascular pathophysiology observed in preeclampsia.

The Reduced Uterine Perfusion Pressure (RUPP) Model of Preeclampsia Causes Decreased Capillary Perfusion in Skeletal Muscle

The Reduced Uterine Perfusion Pressure (RUPP) model of preeclampsia exhibits angiogenic imbalance, endothelial dysfunction, increased vascular resistance and hypertension. We hypothesized the RUPP model would also exhibit decreased functional capillary density and perfusion in skeletal muscle. &#x0D; Female Sprague‐Dawley rats (n=30) were randomized to nonpregnancy (NP) or breeding (Preg) at 12 weeks of age and again to RUPP or SHAM surgeries on gestational day (GD) 14 (or equivalent age in NP rats). On GD 20 (or equivalent), capillary structure and perfusion of the extensor digitorum longus was imaged using digital intravital video microscopy. Functional videos were analyzed by a blinded observer. Capillary density (CD) was expressed as capillaries/millimeter intersecting 3 staggered reference lines (150µm). Flow was scored as the percentage of capillaries having: 1) continuous, 2) intermittent, or 3) stopped flow. &#x0D; Total CD was not different between groups. The RUPP model decreased continuous flow vessels (main effect of surgery, P&lt;0.01) and increased stopped flow (main effect, p&lt;0.01) both which were more pronounced in Preg animals (Continuous: Preg‐SHAM = 80.1±7.8% vs Preg‐RUPP = 67.8±11.2%, p&lt;0.05) (Stopped: Preg‐SHAM = 8.7±3.2% vs Preg‐RUPP = 17.9±5.7% p&lt;0.01). &#x0D; Our results demonstrate that the RUPP model of preeclampsia (and not just general hypoperfusion) is associated with a decreased capillary perfusion in skeletal muscle. &#x0D; Supported by AIHS, Olympus and Q‐Imaging

Development of a High-Pressure, High-Temperature, Optically Accessible Continuous-Flow Vessel for Fuel-Injection Experiments

A vessel has been designed for nonreacting fuel-injection experiments with continuous flow of sweep gas at pressures up to 1380 kPa and temperatures up to 200 °C. Four orthogonal windows provide optical access for high-speed spray-visualization using a fast-pulsed light emitting diode (LED) and a high-speed camera. Initial experiments have been conducted to determine spray characteristics of n-heptane. At room conditions, liquid length and cone angle were 170 mm and 14.5 deg, respectively. With air flow in the chamber at 690 kPa and 100 °C, liquid length was considerably shorter at 92 mm and cone angle was wider at 16.5 deg.

Development and Scale-Up of a Continuous, High-Pressure, Asymmetric Hydrogenation Reaction, Workup, and Isolation

A fully continuous process including an asymmetric hydrogenation reaction operating at 70 bar hydrogen, aqueous extraction, and crystallization was designed, developed, and demonstrated at pilot scale. This paper highlights safety, quality, and throughput advantages of the continuous reaction and separations unit operations. Production of 144 kg of product was accomplished in laboratory fume hoods and a laboratory hydrogenation bunker over two continuous campaigns. Maximum continuous flow vessel size in the lab hoods was 22 L glassware, and maximum plug flow tube reactor (PFR) size in the bunker was 73 L. The main safety advantages of running the hydrogenation reaction continuous rather than batch were that the flow reactor was smaller for the same throughput and, more importantly, the tubular hydrogenation reactor ran 95% liquid filled at steady state. Therefore, the amount of hydrogen in the reactor at any one time was less than that of batch. A two-stage mixed suspension–mixed product removal (MSMPR) cascade was used for continuous crystallization. Impurity rejection by continuous crystallization was superior to that by batch because scalable residence time and steady-state supersaturation enabled robust and repeatable control of enantiomer rejection in a kinetic regime, although this is a nonstandard approach, debatable as an impurity control strategy. The fully continuous wet-end process running in a laboratory infrastructure achieved the same weekly throughput that would be expected from traditional batch processing in a plant module with 400 L vessels.

Identical Feed Drops in a Flow Vessel May not Always Be Desirable

In this paper, the effect of feed size and daughter drop size distributions on the steady-state drop size distribution in a continuous flow vessel is discussed. For identical sizes of feed drops, a new criterion has been proposed to ensure the complete breakage of the feed drops giving rise to smooth continuous drop size distributions. This has been predicted on the basis of two competing time scales, namely, breakage and residence times.

A population balance model for disperse systems: Drop size distribution in emulsion

A mechanistic model for the description and prediction of the drop size distribution in a turbulently agitated emulsion dispersion is developed. Based on this model, breakage rate and relative distribution functions are determined and used to solve the general population balance equation for a continuous flow vessel. Parameters of the models are evaluated from experimental data by means of Marquardt’s method. The model is found to be capable of predicting bimodal drop size distributions of emulsions.

Aggregation and gelation—II. Mixing effects in continuous flow vessels

The state of mixing in a continuous flow vessel is shown to affect the extent of aggregation and the extent of aggregation at which mathematical gelation occurs. Three aspects of the state of mixing in vessels where aggregation alone occurs are considered: the degree of segregation, the residence-time distribution (RTD) and the earliness or lateness of mixing. The effect of the state of mixing is different for each kernel and depends primarily on the order of the moment rate law for the zeroth and sixth moments. For example, the sum kernel has a first-order decay moment rate law for its zeroth moment and is not affected either by the degree of segregation or the earliness or lateness of mixing. On the other hand, the product kernel (ω = 1) has a second-order growth moment rate law for its sixth moment and its gelling behaviour is strongly affected by the degree of segregation and the earliness or lateness of mixing. Our results follow directly from an analogy with reaction engineering based on the formal equivalence of our moment rate law for aggregating systems and a well-known reaction rate law. We show the following striking progression of the dependence of the gelling behaviour of the sum kernel on the RTD: it is a non-gelling kernel in a plug flow vessel, a gelling kernel in a well-mixed vessel and an instantaneously gelling kernel when a vessel contains a partially stagnant zone. We propose that these observations be used for predicting the effects of scale-up and also of departures from ideal mixing. For gelling kernels, any deviation from the ideal case of plug flow always leads to a reduction in th extent of aggregation at which mathematical gelation occurs. Finally, we recommend that batch experiments be used to obtain aggregation rate data because of the difficulty of interpreting unambiguously the data from continuous flow vessels where the state of mixing is not well defined.

Recursive Process Identification Under Variable Flow and Volume

Models of continuous flow vessels with variable volume and flow are studied and methods are developed for their recursive identification. New, composite variables have been introduced recently and it has been proven that many time-variable flow systems have only constant parameters, if they are modelled as functions of such variables, assuming their flow patterns are invariant under changes of flow and volume. The recursive prediction error algorithm is now first formulated in terms of the variable which matches to the case of constant volume and variable flow, and the process is identified by estimating a set of 20 successive, equidistant values of the weighting function. Similar algorithms are then written and identifications executed in the cases of variable volume and flow. The study is extended to the case of a flow pattern which changes independently of the simultaneous variation of the total flow and volume, and the parameters of the pattern are tracked. The identifications of the simulated processes prove the superiority of the approach with regard to algorithms written in terms of the regular time variable. Extension of the method to estimation of chemical reaction parameters is discussed against this background.

Process Control Under Variable Flow and Volume

Models of continuous flow vessels with variable volume and flow are studied and their control algorithms developed. It is shown that the weighting functions of many vessels which are time-variable under such conditions, are invariant with regard to new, composite variables which are introduced. The validity of the basic assumption on invariability of the flow pattern under flow and volume variations is shown valid for several, analytically known cases. The approach is extended to arbitrary, non-analytic response functions obtained by measurements. Automatic controllers are then introduced for processes with variable liquid volume and flow. Both the feedforward and feedback controllers developed in terms of the new variables have constant parameters and eliminate correctly the effects of the disturbance variables in control of concentration or other material property. The corresponding gain scheduling controllers are also presented which operate in terms of the time. Both types of controllers appear easy to implement in hardware and software. Industrial buffer vessels, natural water reservoirs and physiological systems are stated as potential applications of modelling and control, and extensions to chemical reactors are discussed.

Tracer Responses and Control of Vessels with Variable Flow and Volume

Continuous flow vessels which are subject to variation of flow and volume are characterized by time-variable parameters. It is shown that their residence time distributions and weighting functions obtained by tracer testing are made invariant with regard to the integrated flow variables which are introduced. Under variable flow but constant volume, one such integrated variable is sufficient. Under variable volume, two different variables are suggested for the residence time distribution and weighting function, while the appropriate variable of the perfect mixer differs distinctly from that of vessels with a distinct velocity profile. It is shown through a number of example cases, that an agreement with their mathematical models is reached. The approach is extended to include also arbitrary, non-analytic response functions obtained by tracer measurements. Applications of the derived models and their incorporation in automatic control algorithms is discussed.

Variable parameter model of the continuous flow vessel

Models of continuous flow vessels with variable volume and flow are studied. It is shown that the residence distributions and weighting functions of many types of continuous flow vessels which are time-variable under such conditions, are invariant with regard to new, composite variables which are introduced. In addition to the simple, basic flow patterns, the Gaussian distribution of liquid velocity and the laminar flow are in agreement with the presented theory. The approach is extended to arbitrary, non-analytic response functions obtained by measurements. Application of dispersion models to time-variable hydrodynamic processes appears to be subject to limitations. In the case of variable volume, the basic variable of perfect mixer differs distinctly from that of other vessels. Applications and uses of the derived models of variable parameter processes are discussed.

Characterizing the effect of feed rate in continuous flow vessels having turbine agitators

Characterizing the effect of feed rate in continuous flow vessels having turbine agitators

Description of interaction processes in agitated liquid-liquid dispersions

Phenomenological models are proposed to describe drop breakup and coalescence in a turbulently agitated liquid-liquid dispersion. Based on these models, breakage and coalescence rate functions are developed and used to solve the general population balance equation describing drop interactions in a continuous flow vessel. Parameters of the models are evaluated by comparison with experimental data on drop size distributions and mixing frequencies obtained in a continuous flow vessel over a range of operating conditions. The favorable agreement between experimental observation and the model are encouraging that the model is suitable for predicting dispersion properties such as drop size distributions, interfacial areas and mixing frequencies.

The effect of changing viscosity on the transient response of a reaction vessel

The material presented in this paper demonstrates that small continuous flow vessels behave like perfect mixers when mechanical agitation is applied and behave very nearly like perfect mixers even without the use of an agitator. The agitation of the inlet stream seems to give sufficient agitation. The case of a high viscosity fluid replaced by one of lower viscosity follows quite different behavior. For extremely low flow rates the system seems to exhibit almost perfect displacement if there is no agitation but behaves as a perfect mixer if there is agitation. With moderate flow rates the behavior is similar with or without an agitator and shows characteristics of dead space, channeling and diffusion.