Method and system for testing component content of hydroxylammonium nitrate-based propellant

Abstract

The invention discloses a hydroxylammonium nitrate-based propellant component content testing method and system, and the method comprises the steps: (1) obtaining a standard sample set, the standard sample set comprises a plurality of standard samples, each standard sample comprises a plurality of components of a hydroxylammonium nitrate-based propellant, and the content of each component is known; (2) collecting the near infrared spectrum of the standard sample set to form a standard spectrum data set; (3) on the basis of the standard spectrum data set and the known content of each component, constructing a quantitative analysis model for predicting the content of each component through a multivariate correction algorithm; and (4) collecting the near infrared spectrum of a to-be-detected sample, and simultaneously predicting the contents of various components in the to-be-detected sample by using the quantitative analysis model. The system comprises a near-infrared spectrometer, a temperature control device and a calculation processing unit. The testing method provided by the invention can realize one-step determination of the content of all components, and is simple and convenient to operate, rapid, non-destructive, accurate and reliable, so that an efficient quality assurance means is provided.

Description

Technical Field

[0001] This invention belongs to the field of chemical testing, specifically, it relates to a method and system for testing the content of components in hydroxylamine nitrate-based propellants. More specifically, it relates to a rapid, non-destructive, and simultaneous testing method and system for the content of multiple components in hydroxylamine nitrate-based propellants based on near-infrared spectroscopy and chemometrics. Background Technology

[0002] With the completion of my country's space station and the proposal of its manned lunar landing program, the development of manned spaceflight has entered a new stage. Achieving non-toxic propulsion systems for manned spacecraft is a necessity for the development of manned spaceflight, and a crucial approach to this is the use of non-toxic propellants. Hydroxyammonium nitrate-based propellants (HAN-based propellants) are a novel type of ionic liquid propellant composed of hydroxyammonium nitrate, compatible fuels, additives, and water. They possess characteristics such as being green and pollution-free, having adjustable specific impulse, high density, and low saturated vapor pressure, and have become a key research focus for controllable propulsion sources for spacecraft.

[0003] A typical HAN-based propellant is a homogeneous aqueous solution composed of multiple components, including the oxidant hydroxyl ammonium nitrate (HAN), fuel (usually a water-soluble organic alcohol or amine), high-energy additives (such as nitrates), co-solvents, and water. Its final properties, such as specific impulse, combustion temperature, catalytic decomposition characteristics, and storage stability, are strictly dependent on the precise proportions of each component. Deviations in the content of any component from the design range can have a significant, even fatal, adverse effect on the overall performance of the propellant. Therefore, establishing rapid, accurate, and reliable methods for analyzing component content is crucial for ensuring the product quality, batch consistency, and safe use of HAN-based propellants.

[0004] Currently, the determination of the total component content of HAN-based propellants typically relies on a combination of traditional wet chemical analysis methods. For example, the content of hydroxylamine nitrate may be determined using a specific titration method; certain organic additives may require separation and detection by gas chromatography or liquid chromatography; while the water content is often obtained using the Karl Fischer method or a simple difference method. This "divide and conquer" analytical strategy has significant limitations: First, it requires the use of multiple different analytical instruments and chemical reagents, resulting in a cumbersome process, long analysis cycle, and high cost; second, multiple sampling and the introduction of multiple methods inevitably increase the accumulation of systematic and random errors, affecting the accuracy of the final results; third, some methods are destructive analyses, making it impossible to perform non-destructive testing on valuable samples; and finally, the complex operating procedures require highly skilled analysts, making it difficult to meet the needs of rapid, high-volume quality monitoring on production lines.

[0005] Near-infrared spectroscopy (NIRS) is a rapidly developing green analytical technique. It combines molecular vibrational spectroscopy (primarily the overtone and combination frequency absorptions of hydrogen-containing groups such as CH, OH, and NH) with modern chemometric methods, enabling simultaneous, rapid, and non-destructive quantitative analysis of multiple components in complex mixtures. The basic process involves collecting the spectra of a representative set of standard samples and using their known reference content values ​​to establish a mathematical relationship (i.e., a quantitative model) between the spectral data and the content through a multivariate correction algorithm (such as partial least squares PLS). Subsequently, for unknown samples, simply measuring their near-infrared spectra allows for rapid prediction of the content of each component using the established model.

[0006] However, successfully applying NIRS technology to the specific system of HAN-based propellants faces several unique technical challenges: 1. Interference from Free Nitric Acid: The key raw material for HAN production—an aqueous solution of hydroxyl ammonium nitrate—usually contains a small amount of unreacted free nitric acid during the preparation process, and its content may fluctuate randomly between different production batches. This "unplanned" nitric acid is introduced into the final propellant product, and its near-infrared absorption signal may overlap or interfere with the signals of the main components (especially HAN and water). If this is not considered in the modeling, it will lead to systematic biases in the prediction model during practical application.

[0007] 2. Temperature Sensitivity: Near-infrared spectroscopy is highly sensitive to temperature changes. Temperature variations alter hydrogen bond strength and intermolecular forces, leading to changes in the position, intensity, and shape of absorption peaks. Laboratory temperatures inevitably fluctuate; if the quantitative model is not robust to temperature changes, it will severely impact the accuracy and stability of the prediction results.

[0008] 3. Inter-component interactions: HAN-based propellants are multi-component aqueous solutions with high ionic strength. There are complex interactions between the components (such as hydrogen bonding and ion association), which may make the spectrum-concentration relationship not a simple linear superposition, increasing the complexity of modeling.

[0009] 4. Signal Saturation and Band Selection: Water exhibits strong absorption in the near-infrared region, which may lead to signal saturation and loss of useful information in certain bands. Therefore, it is essential to carefully select the spectral range used for modeling, avoiding saturation regions, and identifying the characteristic bands most relevant to the content of each component.

[0010] In summary, there is an urgent need in this field to develop a component content testing method that can overcome the above challenges, is specifically designed for the characteristics of nitrate hydroxyl ammonium-based propellants, and integrates speed, accuracy, non-destructive testing, and high efficiency.

[0011] In view of this, the present invention is proposed. Summary of the Invention

[0012] The technical problem to be solved by this invention is to overcome at least one of the shortcomings of the prior art and provide a method and system for testing the content of hydroxylamine nitrate propellant components. This testing method aims to achieve the goal of simultaneously obtaining the content of all major components in a single measurement, and has advantages such as simple pretreatment, convenient operation, fast analysis speed, non-destructive testing, and accurate and reliable results.

[0013] To solve the above-mentioned technical problems, the basic concept of the technical solution adopted by the present invention is as follows: In a first aspect, the present invention provides a method for testing the content of nitrate hydroxylamine-based propellant components, comprising the following steps: (1) Obtain a standard sample set, which contains multiple standard samples, each of which contains multiple components of nitrate hydroxyl ammonium propellant, and the content of each component is known; (2) Collect the near-infrared spectra of the standard sample set to form a standard spectral dataset; (3) Based on the standard spectral dataset and the known content of each component, a quantitative analysis model for predicting the content of each component is constructed using a multivariate calibration algorithm; (4) Collect the near-infrared spectrum of the sample to be tested, and use the quantitative analysis model to predict the content of multiple components in the sample to be tested.

[0014] Compared to combined determination methods using multiple testing techniques for different components, this invention utilizes a molecular spectral multivariate correction method to establish component content models. This allows for separate analysis of each component's content using different component models on the spectrum of a single sample measurement, achieving one-step determination of the entire component content. This solves the analytical bottleneck problem of large cumulative errors from multiple tests in HAN-based propellants, significantly improving analytical efficiency and resulting in substantial cost reduction and efficiency gains. Thus, the method of this invention enables rapid, synchronous, non-destructive, and accurate determination of the multi-component content of the sample. The method features simple pretreatment and convenient operation, significantly improving analytical efficiency and reliability, and providing an efficient quality assurance means for propellant production and application.

[0015] In a further scheme, step (1) involves designing a standard sample set by: designing the content of hydroxyl nitrate, additives, and water in the standard sample set based on the chemical composition and technical specifications of the hydroxyl nitrate-based propellant, wherein the additives may be one or more depending on the specific chemical composition of the propellant; the variation range of the above components in the sample set at least covers the product technical specifications range, or extends appropriately beyond the technical specifications range.

[0016] In a further scheme, in step (1), the method for preparing the standard sample set includes: using hydroxyl ammonium nitrate, additives, water, and / or a mixture containing two or more components contained in the propellant as raw materials, mixing them in a certain proportion according to the designed sample composition to prepare the standard sample set, and determining the accurate values ​​of hydroxyl ammonium nitrate, additives, and water in the standard sample set according to the composition of the raw materials and the amount of materials added during preparation.

[0017] In a further step, the near-infrared spectra of the standard sample set in step (2) can be acquired using one of the following methods: a) Directly collect the near-infrared spectra of the prepared standard sample set and establish a quantitative analysis model; b) Add a quantitative amount of nitric acid or nitric acid aqueous solution to some or all of the standard samples, mix thoroughly, collect their near-infrared spectra, and establish a quantitative analysis model; c) First, collect the near-infrared spectra of the original standard sample set. Then, add a certain amount of nitric acid or nitric acid aqueous solution to some or all of the standard samples, mix them thoroughly, and collect their near-infrared spectra again. Use the near-infrared spectra of the standard sample set before and after adding nitric acid to establish a quantitative analysis model.

[0018] Hydroxylammonium nitrate solution is a key raw material for the production of hydroxylammonium nitrate-based propellants. It is prepared by neutralizing nitric acid and hydroxylamine solutions. Due to requirements regarding production processes, safety, and stability, a small amount of nitric acid is usually added beyond the stoichiometric ratio during the preparation of the hydroxylammonium nitrate solution. This results in the presence of a small amount of unneutralized free nitric acid in the solution. This free nitric acid forms excess nitric acid in the newly produced hydroxylammonium nitrate-based propellant, present in a randomly distributed manner outside the stoichiometric composition, and causes changes in the near-infrared spectrum of the propellant.

[0019] Unlike actual products where the excess nitric acid content and ammonium nitrate content in propellant products prepared using different batches of ammonium nitrate solution exhibit a random distribution, standard sample sets are typically prepared using the same batch of ammonium nitrate solution. This results in a strong linear correlation between the excess nitric acid content and the ammonium nitrate content in the standard sample sets; or, because the ammonium nitrate solution used to prepare the standard sample sets does not contain free nitric acid, the resulting standard sample sets are free of excess nitric acid. The near-infrared spectral model established using this standard sample set cannot cover actual propellant products, especially those with high ammonium nitrate content, where the excess nitric acid content varies randomly. This means that during actual product testing, the excess nitric acid may interfere with the content analysis of other components.

[0020] Therefore, this invention directly acquires the near-infrared spectra of the prepared standard sample set and establishes a model; or, within a certain addition range, different masses of nitric acid / nitric acid aqueous solution are added to some or all of the standard samples, mixed thoroughly, and then their near-infrared spectra are acquired and a model is established; or, near-infrared spectra are first acquired for the prepared standard sample set, and then, within a certain addition range, different masses of nitric acid / nitric acid aqueous solution are added to some or all of the standard samples, mixed thoroughly, and then the near-infrared spectra of the standard samples with added nitric acid / nitric acid aqueous solution are acquired again. The near-infrared spectra of the standard sample set before and after the addition of nitric acid are combined to establish the model. This more closely reflects the actual situation of real products having excess nitric acid with random distribution, improving the accuracy of the prediction results. During the model establishment process, this portion of nitric acid is identified by the multivariate correction algorithm and distinguished from other major components in the propellant, thereby avoiding interference with the content analysis of the major components.

[0021] Specifically: Method a (direct collection) serves as the basic approach, with a simple and efficient process. It is suitable for establishing benchmark models or for rapid analysis under ideal conditions where the free nitric acid content in the known raw materials is extremely low and stable.

[0022] Secondly, Method b (data collection after addition) and Method c (data collection combined before and after addition) are innovative solutions to address practical engineering challenges. During the production of ammonium nitrate solution, fluctuating amounts of free nitric acid are typically introduced, resulting in unplanned nitric acid content in the final product. This randomly distributed free nitric acid alters the near-infrared spectral characteristics of the system. If the model fails to learn this variation, significant errors will occur in actual predictions, especially affecting the determination of ammonium nitrate and water content. Method b, by actively introducing different amounts of nitric acid into standard samples, simulates the fluctuations of free nitric acid in real products, enabling the established model to learn and "recognize" the spectral characteristics of nitric acid. Method c goes further, not only including the spectrum in the presence of nitric acid but also retaining the spectrum of the original sample, providing the model with richer and more continuous spectral variation information. This allows the model not only to identify nitric acid but also to better analyze the interaction between nitric acid and other major components (such as ammonium nitrate, water, and additives), thus more accurately separating the nitric acid signal from overlapping peaks during prediction, effectively avoiding its interference with the quantitative results of the target component. Preferably, the method of collecting near-infrared spectra of standard sample sets is selected as c.

[0023] Thus, this invention can fundamentally solve the problem of inaccurate prediction caused by the random distribution of free nitric acid, enabling the near-infrared analysis method to be widely applied to the quality control of real products from different batches and sources, with accurate and reliable results.

[0024] A further approach involves calculating the amount of nitric acid / nitric acid aqueous solution to be added based on the content of free nitric acid in the ammonium nitrate solution and the content of ammonium nitrate in the HAN-based propellant, so that the mass fraction of nitric acid in the standard sample after addition is between 0% and 0.5%.

[0025] Temperature can indirectly affect the accuracy and stability of multivariate calibration models by altering the spectral characteristics of the analyte, such as the position and intensity of absorption peaks, potentially leading to model failure or large prediction errors. Therefore, this invention incorporates temperature variation into the spectral acquisition process.

[0026] In a further step, during steps (2) and (4), the temperature control in the near-infrared spectral acquisition process shall employ one of the following methods: a) All standard samples and test samples were spectrally acquired at the same temperature; b) Within the set temperature range, select multiple different temperature conditions to collect spectra of the standard samples, and select one or more temperature conditions to collect spectra of each standard sample; the sample to be tested is collected at any temperature within the set temperature range.

[0027] In this invention, during spectral acquisition, all standard samples are controlled to undergo spectral acquisition at the same temperature, and the spectral acquisition of the sample to be tested is also performed at that temperature to ensure the accuracy of the prediction results. Alternatively, spectral acquisition of standard samples can be performed under multiple different temperature conditions within a certain temperature range. Each standard sample can select one or more temperature conditions for spectral acquisition, making the model more temperature adaptable, reducing the need for temperature control capabilities of the near-infrared spectrometer, minimizing the influence of ambient temperature on the test results, and improving the accuracy of the test results. When testing the sample to be tested, spectral acquisition can be performed at any temperature within this temperature range. Specifically: The samples in Method a are all acquired at the same constant temperature, which can eliminate the spectral differences introduced by temperature variables to the greatest extent, providing a highly consistent thermodynamic environment for model building and prediction, and ensuring the accuracy and reproducibility of the model under ideal conditions.

[0028] Method b incorporates temperature variations into the model training beforehand, enabling the established quantitative analysis model to inherently possess the ability to compensate for and correct for the effects of temperature. This significantly enhances the model's adaptability, reduces the stringent requirements for constant temperature conditions during prediction, and allows it to maintain stable and reliable prediction results in real-world environments. This makes the analysis method adaptable to a wider range of field application scenarios.

[0029] As a preferred embodiment, during the near-infrared spectral acquisition process of the sample in this invention, temperature control adopts method b).

[0030] In a further scheme, in steps (2) and (4), when collecting near-infrared spectra, the temperature of the standard sample and the sample to be tested is controlled between 5℃ and 40℃, covering almost all possible laboratory temperatures.

[0031] In a further step, in step (3), the standard spectral dataset is first preprocessed, and then the preprocessed near-infrared spectral data containing absorbance at multiple wavelengths is correlated with the content values ​​of each component through a multivariate correction algorithm to construct a quantitative analysis model.

[0032] During sample acquisition, the sample's state, color, and instrument response often introduce interference from factors unrelated to the sample's properties into the raw near-infrared spectrum. This leads to baseline drift and spectral instability, making preprocessing essential. Preprocessing involves performing a series of mathematical transformations and processing on the raw near-infrared spectral data before building a quantitative model. Its purpose is to minimize or eliminate interference unrelated to the analyte's content, thereby highlighting effective spectral features relevant to the content and laying the foundation for a stable and accurate quantitative analysis model.

[0033] Hydroxylammonium nitrate propellant is a colorless, transparent, and homogeneous liquid. Therefore, this invention selects one or more commonly used liquid pretreatment methods for comparison. In a further embodiment, the pretreatment method is selected from one or more of the following: baseline correction, standardization, mean centering, first derivative, second derivative, multivariate scattering correction, standard normal variable transformation, detrending, and Savitsky-Golay smoothing.

[0034] For example, preprocessing mainly includes one or more combinations of the following: eliminating baseline drift and tilt (which can be done by baseline correction and detrending), reducing noise interference (which can be done by Savitzky-Golay smoothing), enhancing spectral feature differences (which can be done by first derivative and second derivative), correcting scattering effects (which can be done by multivariate scattering correction and standard normal variable transformation), and standardizing data scales (which can be done by standardization and mean centering).

[0035] In a further embodiment, the multivariate correction algorithm is selected from one or more of the following: multivariate linear regression, principal component regression, and partial least squares.

[0036] When establishing the spectral model, the multivariate correction method is used. When modeling with absorbance data at a small number of discrete wavelengths, multivariate linear regression is used. When modeling with wavelength ranges containing a large number of wavelengths, principal component regression or partial least squares is used.

[0037] In a further step, in step (3), when establishing the quantitative analysis model, the spectral data used is selected from one of the following methods: a) Use absorbance data across all wavelengths collected; b) Select absorbance data within one or more wavelength ranges; c) Select absorbance data at multiple discrete characteristic wavelength points.

[0038] Method a uses all wavelength data, which can make the most of all information in the spectrum and avoid missing some weak but important characteristic peaks due to subjective selection.

[0039] Method b selects one or more characteristic wavelength ranges. By eliminating irrelevant or noisy bands, the interference of irrelevant variables can be reduced, allowing the model to focus more on the spectral information most relevant to the target component. This can speed up the calculation. The selected wavelength range corresponds to the combination and overtone absorption regions of specific molecular vibrations of the component to be measured (such as the NH bond of ammonium nitrate, the OH bond of water, and the CH bond of auxiliaries), which can enhance the interpretability of the model.

[0040] Method c selects multiple discrete characteristic wavelength points, resulting in the simplest model, strong resistance to overfitting, and high computational efficiency.

[0041] A further approach is to exclude spectral data in the wavelength range of 1460 nm to 1660 nm when establishing a quantitative analysis model.

[0042] In near-infrared spectroscopy, the 1460 nm to 1660 nm range represents a strong absorption region of the first harmonic of water (the first harmonic of the stretching vibration of the OH bond). For aqueous solutions of hydroxylamine nitrate propellants, the high concentration of water leads to extremely strong absorbance in this band, even exceeding the linear range of the detector, resulting in signal saturation. Saturated signals do not contain effective concentration gradient information and only introduce significant noise and nonlinear errors. Furthermore, the signal in the saturated region is unstable, amplifying noise. Excluding these "invalid" or even "harmful" data from the model can significantly improve the signal-to-noise ratio (SNR) of the entire model. This allows the algorithm to more clearly "see" the characteristic absorptions of other components (hydroxylamine nitrate, additives) in other bands, thereby establishing more accurate and reliable quantitative relationships. Therefore, excluding spectral data in the 1460 nm to 1660 nm wavelength range avoids saturation interference from strong water absorption, improves the SNR and model accuracy, and enhances the robustness of the model.

[0043] A further approach involves continuously or intermittently purging the sample chamber of the spectrometer with dry gas before, during, and after spectral acquisition to prevent moisture condensation on the surface of optical components and to avoid fogging on the cuvette surface caused by environmental humidity, which would affect the accuracy of the test results.

[0044] In a further step, in step (4), after obtaining the contents of ammonium nitrate and all auxiliaries using a quantitative analysis model, the water content is calculated by the difference method. A further proposed solution is that the water content = 100% - (the content of ammonium nitrate + the sum of the contents of all additives).

[0045] Secondly, the present invention provides a near-infrared spectroscopy analysis system for implementing the aforementioned test method, the system comprising: Near-infrared spectrometer, used to collect near-infrared spectra of standard samples and samples to be tested; Temperature control device, used to adjust and stabilize the temperature of the sample during spectral acquisition; The computing processing unit is configured as follows: Store the standard spectral dataset and the known content of each component; A multivariate calibration algorithm is executed to establish the quantitative analysis model. The spectrum of the sample to be tested is received, and the content of each component is calculated and output using the quantitative analysis model.

[0046] Thirdly, the present invention provides a method for constructing a standard sample set for quantitative analysis of hydroxylamine nitrate propellants, the method comprising: Based on the chemical composition and technical specifications of the hydroxyl ammonium nitrate propellant, the design concentration ranges of hydroxyl ammonium nitrate, various additives, and water are determined. Using ammonium nitrate, additives, water, and / or mixtures containing two or more propellant components as raw materials, multiple standard samples covering the designed concentration range are prepared by precise weighing and mixing. Preferably, different amounts of nitric acid or nitric acid aqueous solution are precisely added to some or all of the prepared standard samples to simulate the random distribution of free nitric acid in the actual product, thus forming an extended standard sample set.

[0047] Fourthly, the present invention provides a non-transitory computer-readable storage medium storing a computer program configured, when executed by a processor, to implement the method for testing the content of the nitrate hydroxyl ammonium-based propellant component.

[0048] By adopting the above technical solution, the present invention has the following beneficial effects compared with the prior art.

[0049] (1) High efficiency and high throughput: This invention establishes a method for analyzing the components of hydroxylamine nitrate propellants based on molecular spectroscopy multivariate correction technology, achieving "one-step determination of all component contents and simultaneous output of all components." This shortens the multi-step analysis process, which originally required several hours or even longer, to within minutes, significantly improving analytical efficiency. It is particularly suitable for rapid quality control and large-scale sample screening in the production process. In practical work, it is only necessary to scan the near-infrared spectrum of the sample to be tested and substitute it into the model to simultaneously predict the contents of hydroxylamine nitrate and various additives. Compared with the requirement of multiple testing methods and multi-step testing in relevant specifications, the near-infrared spectroscopy method is simple to preprocess, easy to operate, fast and non-destructive, accurate and reliable, providing an efficient and reliable quality assurance means for propellant production and application.

[0050] (2) The test method developed in this invention adds a small amount of nitric acid to the standard sample. Since the raw material, ammonium nitrate solution, contains an unpredictable amount of free nitric acid, the actual product contains excess nitric acid with a random distribution within a certain range, which is outside the quantifiable composition. Therefore, this invention is more consistent with reality and improves the accuracy of the prediction results. Thus, by introducing the "nitric acid addition" strategy, the model is able to resist interference from free nitric acid in the actual product, fundamentally solving the prediction deviation problem caused by batch differences in raw materials, especially significantly improving the accuracy of the determination of high HAN content propellants.

[0051] (3) The testing method developed in this invention incorporates sample spectra at different temperatures within the range of 5℃-40℃ into the model during spectral acquisition by using variable-temperature scanning. This covers almost all possible laboratory temperatures, making the model more temperature adaptable, reducing the need for temperature control capabilities of the near-infrared spectrometer, and ensuring the accuracy of test results under varying temperature conditions. Thus, the "variable-temperature modeling" strategy greatly enhances the model's environmental adaptability (temperature robustness), ensuring the stability and comparability of test results in different seasons and laboratory environments.

[0052] (4) The testing method developed in this invention can perform compositional analysis of hydroxyammonium nitrate propellants without altering the sample composition, demonstrating its potential for in-situ, online propellant detection. Furthermore, it is free of toxic and harmful chemical reagents, ensuring the safety of testing personnel. The entire analytical process requires no toxic or harmful chemical reagents and poses no threat to the environment or the health of operators. Moreover, this method is non-destructive, and the measured sample can still be used for other purposes, making it significant for studying valuable samples or conducting long-term stability tracking studies.

[0053] (5) Near-infrared spectrometers are becoming increasingly mature and widespread. Once a model is created, it can be used repeatedly for a long time, and the cost of a single test is extremely low. This method is simple to operate, requires relatively low personnel skills, and is easily integrated with automated sample introduction systems and online processes, laying a solid technical foundation for realizing real-time online quality monitoring of the HAN-based propellant production process.

[0054] The specific embodiments of the present invention will now be described in further detail with reference to the accompanying drawings. Attached Figure Description

[0055] The accompanying drawings, as part of this invention, are used to provide a further understanding of the invention. The illustrative embodiments and descriptions of the invention are used to explain the invention, but do not constitute an undue limitation of the invention. Obviously, the drawings described below are merely some embodiments, and those skilled in the art can obtain other drawings based on these drawings without creative effort. In the drawings: Figure 1 This is the near-infrared spectrum of a sample set of hydroxylamine nitrate propellant.

[0056] It should be noted that these accompanying drawings and textual descriptions are not intended to limit the scope of the invention in any way, but rather to illustrate the concept of the invention to those skilled in the art by referring to specific embodiments. Detailed Implementation

[0057] To make the objectives, technical solutions, and advantages of the embodiments of the present invention clearer, the technical solutions in the embodiments will be clearly and completely described below with reference to the accompanying drawings. The following embodiments are used to illustrate the present invention, but are not intended to limit the scope of the present invention.

[0058] Example 1 (1) Design of standard sample set The specific design of standard sample set A in this embodiment is as follows: the standard sample set is designed with reference to the technical specifications of type I HAN-based propellant, covering the concentration range of sample sets for ammonium nitrate, alcohols, high-energy additives and cosolvents, and appropriately extended outward based on the coverage of the technical specifications range.

[0059] Table 1. Component Concentration Design of Standard Sample Set (2) Preparation of standard sample set Using aqueous solutions of ammonium nitrate, pure alcohols, high-energy additives, and co-solvents as raw materials, the accurate content of ammonium nitrate and additives in each raw material was determined in advance using a traceable standard method. Based on the designed sample composition and the concentration of each raw material, the raw materials were mixed in a certain proportion to prepare a standard sample set. The accurate values ​​of ammonium nitrate and each additive in the standard sample set were determined according to the composition of the raw materials and the amount of materials used during preparation, resulting in standard sample set A.

[0060] Preparation process: According to the design table, accurately weigh each raw material using a 0.01% analytical balance and place them in a stoppered glass bottle. Ensure uniform mixing by magnetic stirring or vortex oscillation to form approximately 50-100 standard samples covering the entire design space. Record the precise amount of each sample and calculate its accurate reference content of components.

[0061] Example 2 This embodiment is based on the standard sample set prepared in Example 1, and establishes a basic model (model group A) for collecting spectra at a constant temperature. The specific modeling steps for model group A are as follows: (1) Spectral acquisition: Instrument: Grating type near-infrared spectrometer.

[0062] Parameters: Scan range 1000-1800 nm.

[0063] Operation and conditions: Scan the standard sample set A in Example 1, scanning each standard sample once, with the sample temperature controlled at 20℃±0.2℃. Before, during, and after spectral acquisition, continuously purge the sample chamber with dry nitrogen gas to ensure that no fog is generated on the surface of the quartz cuvette.

[0064] Near-infrared spectra of the hydroxylamine nitrate propellant sample set are as follows: Figure 1 As shown.

[0065] (2) Establishment of quantitative model: Preprocessing: The original spectrum is processed by "mean centering" and "first derivative" to eliminate baseline drift and enhance spectral features.

[0066] Band selection: To avoid signal saturation caused by strong water absorption, spectral data from two bands, 1000-1450 nm and 1670-1800 nm, were selected for modeling.

[0067] Modeling Algorithm: The selected near-infrared spectral data were fitted to the component contents of each sample in the standard sample set A. Partial least squares (PLS) was used to establish independent quantitative models for ammonium nitrate, alcohols, high-energy auxiliaries, and cosolvents. Leave-one-out cross-validation was used to determine the optimal number of principal factors for each model, and the root mean square error (RMSECV) was used as the evaluation index for model performance.

[0068] Example 3 This embodiment establishes a quantitative analysis model (model group B) with enhanced temperature adaptability based on the standard sample set prepared in Example 1. The specific modeling steps for model group B are as follows: (1) Spectral acquisition: Instrument: Grating type near-infrared spectrometer.

[0069] Parameters: The scanning spectral range is 1000~1800 nm.

[0070] Operation and conditions: Scan the standard sample set A in Example 1, scanning each sample three times, with the temperature controlled at 5℃, 20℃, and 40℃ respectively. Before, during, and after spectral acquisition, continuously purge the sample chamber with dry nitrogen gas to ensure that no fog is generated on the surface of the quartz cuvette.

[0071] (2) Establishment of quantitative model: Preprocessing: The original spectrum is processed by "mean centering" and "first derivative" to eliminate baseline drift and enhance spectral features.

[0072] Band selection: To avoid signal saturation caused by strong water absorption, spectral data from two bands, 1000-1450 nm and 1670-1800 nm, were selected for modeling.

[0073] Modeling Algorithm: The selected near-infrared spectral data were fitted with the component contents of each sample in the standard sample set A. This involved correlating the spectral data of each sample at three temperatures with its unique component reference content value. Partial least squares (PLS) was used to establish independent quantitative models for ammonium nitrate, alcohols, high-energy additives, and co-solvents. Leave-one-out cross-validation was employed to determine the optimal number of principal factors for each model, and root mean square error (RMSECV) was used as the performance evaluation metric for the models.

[0074] Example 4 Model group A and model group B were tested using actual product samples of type I HAN-based propellant. Test samples: Two different batches of Type I HAN-based propellant actual products (Product 1 and Product 2) were selected.

[0075] Among them, Product I, batch number HAN-I-2024-03, was produced by Beijing Aerospace Test Technology Research Institute; Product I, batch number 2, is HAN-I-2025-01 and was produced by the Beijing Aerospace Test Technology Research Institute.

[0076] Reference value determination: The content of each component in the two products is accurately determined using the multi-step titration method specified in the relevant national standards or national military standards, and used as a reference value.

[0077] Near-infrared detection: Near-infrared spectra of two actual Type I HAN-based propellant samples were collected by scanning them at 10℃, 20℃ and 30℃ respectively.

[0078] Results Comparison: The above spectra were predicted using model group A (single-temperature model) established in Example 2 and model group B (multi-temperature model) established in Example 3, respectively, and the prediction results were compared with the titration reference values. The test results are shown in Table 2.

[0079] Table 2 Results analysis: As can be seen from the data in Table 2, compared with the results of collecting and establishing the calibration model only at 20℃, model group B, which collected and established the calibration model under multiple temperature conditions, has more accurate prediction results at several different temperatures tested.

[0080] Model group A, established only at 20℃, showed a significant increase in the deviation between its predicted results and the reference values ​​when the test temperature deviated from 20℃ (e.g., 10℃ and 30℃). In contrast, model group B showed a high degree of agreement with the reference values ​​for all components at all test temperatures (10℃, 20℃, 30℃), with very small fluctuations, demonstrating that model group B has superior temperature adaptability.

[0081] Example 5 This embodiment establishes a quantitative analysis model (model group C) that can effectively resist interference from free nitric acid based on the standard sample set prepared in Example 1. The specific modeling steps for model group C are as follows: (1) Spectral acquisition: Instrument: Grating type near-infrared spectrometer.

[0082] Parameters: Scan range 1000-1800 nm.

[0083] Operation and conditions: Scan the standard sample set A in Example 1, scanning each sample once, and maintain the temperature at 20°C. During the measurement, nitrogen gas is introduced into the test chamber to ensure that no fog is generated on the surface of the quartz cuvette.

[0084] (2) Add nitric acid to the sample set: Accurately weigh the remaining mass of each sample in standard sample set A. Add different volumes (e.g., 0.01 ml, 0.02 ml, 0.03 ml, 0.04 ml, and 0.05 ml) of 67% (w / w) nitric acid aqueous solution sequentially to each standard sample. Calculate the new mass fraction of nitric acid in the sample after each addition (ensuring it is within the range of 0–0.5%) and the new content of each component after dilution. Mix thoroughly after each addition of nitric acid.

[0085] (3) Second spectral acquisition: The standard sample with added nitric acid was scanned in (2) under the same conditions as in (1).

[0086] (4) Establishment of quantitative models: Using the same preprocessing methods (mean centering, first derivative) and band selection (spectral range 1000~1450, 1670~1800 nm) as in Example 2, the near-infrared spectral data collected twice were fitted with the component contents of each sample before and after the addition of nitric acid, and a model was established using partial least squares. In this model, the variation of nitric acid was treated as a hidden interference factor and automatically handled by the algorithm.

[0087] Example 6 Model group A and model group C were tested using actual product samples of type I HAN-based propellant. Test samples: In addition to the two Type I HAN-based propellant products (with medium HAN content) used in Example 4, two additional Type II HAN-based propellant products (Product 1 and Product 2) with higher HAN content were selected.

[0088] Among them, the batch number of Type II product 1 is HAN-II-2024-02, which was produced by Beijing Aerospace Test Technology Research Institute; Product II, batch number HAN-II-2024-05, was produced by the Beijing Aerospace Test Technology Research Institute.

[0089] Reference value determination: The content of hydroxylamine nitrate in all products was determined by titration and used as a reference value.

[0090] Near-infrared detection: The spectra of all products were collected at 20°C.

[0091] Results Comparison: The content of hydroxylamine nitrate in the two Type I HAN-based propellant products and the two Type II HAN-based propellant products in Example 4 was predicted using model group A (without considering nitric acid interference) and model group C (considering nitric acid interference), respectively. The results were compared with the reference values ​​of titration method. The test results are shown in Table 3.

[0092] Table 3 sample Titration test results (%) Model A prediction results (%) Model C prediction results (%) Type I HAN-based propellant product 1 36.94 36.91 36.96 Type I HAN-based propellant product 2 37.26 37.19 37.22 Type II HAN-based propellant product 1 62.76 61.94 62.65 Type II HAN-based propellant product 2 65.28 64.46 65.47 Results analysis: As can be seen from the data in Table 3, for Type I products with moderate HAN content, the prediction accuracy of Model A and Model C is comparable. For Type II products with higher HAN content, the predicted values ​​of Model A are systematically lower and have a larger bias; while the predicted values ​​of Model C are very close to the titration reference values.

[0093] These results indicate that the near-infrared spectral model set C, which includes both standard samples with and without added nitric acid, can obtain relatively accurate predictions of nitric acid phosphate content in HAN-based propellants. In contrast, the near-infrared spectral model set, which only uses standard samples without added nitric acid, may produce inaccurate measurement results for certain types of HAN-based propellants, especially those with high HAN content, because the model cannot separate the signal interference from excess nitric acid.

[0094] This invention introduces nitric acid variation during modeling, and model group C successfully learns and compensates for the interference of free nitric acid, significantly improving the prediction accuracy for actual products, especially products with high oxidant content.

[0095] The above description is merely a preferred embodiment of the present invention and is not intended to limit the present invention in any way. Although the present invention has been disclosed above with reference to preferred embodiments, it is not intended to limit the present invention. Any person skilled in the art can make some modifications or alterations to the above-described technical content to create equivalent embodiments without departing from the scope of the present invention. Any simple modifications, equivalent changes, and alterations made to the above embodiments based on the technical essence of the present invention without departing from the scope of the present invention shall still fall within the scope of the present invention.

Claims

1. A method for testing the content of hydroxylamine nitrate-based propellant components, characterized in that, Includes the following steps: (1) Obtain a standard sample set, which contains multiple standard samples, each of which contains multiple components of nitrate hydroxyl ammonium propellant, and the content of each component is known; (2) Collect the near-infrared spectra of the standard sample set to form a standard spectral dataset; (3) Based on the standard spectral dataset and the known content of each component, a quantitative analysis model for predicting the content of each component is constructed using a multivariate calibration algorithm; (4) Collect the near-infrared spectrum of the sample to be tested, and use the quantitative analysis model to predict the content of multiple components in the sample to be tested.

2. The test method according to claim 1, characterized in that, In step (2), the near-infrared spectra of the standard sample set are collected using one of the following methods: a) Directly collect the near-infrared spectra of the prepared standard sample set and establish a quantitative analysis model; b) Add unequal amounts of nitric acid or nitric acid aqueous solution to some or all of the standard samples, mix thoroughly, collect their near-infrared spectra, and establish a quantitative analysis model; c) First, collect the near-infrared spectra of the original standard sample set. Then, add unequal amounts of nitric acid or nitric acid aqueous solution to some or all of the standard samples, mix them thoroughly, and collect their near-infrared spectra again. Use the near-infrared spectra of the standard sample set before and after adding nitric acid to establish a quantitative analysis model. Preferably, the mass of the added nitric acid or nitric acid aqueous solution is controlled so that the final mass fraction of nitric acid in the standard sample is 0%-0.5%.

3. The test method according to claim 1 or 2, characterized in that, In steps (2) and (4), during the near-infrared spectral acquisition process, temperature control employs one of the following methods: a) All standard samples and test samples were spectrally acquired at the same temperature; b) Within the set temperature range, select multiple different temperature conditions to collect spectra of the standard samples, and select one or more temperature conditions to collect spectra for each standard sample; The sample to be tested is subjected to spectral acquisition at any temperature within the set temperature range; Preferably, in steps (2) and (4), when collecting near-infrared spectra, the temperature of the standard sample and the sample to be tested is controlled to be between 5℃ and 40℃.

4. The test method according to any one of claims 1-3, characterized in that, In step (3), the standard spectral dataset is first preprocessed, and then the preprocessed near-infrared spectral data containing absorbance at multiple wavelengths is correlated with the content values ​​of each component through a multivariate correction algorithm to construct a quantitative analysis model. Preferably, the preprocessing method is selected from one or more of the following: baseline correction, standardization, mean centering, first derivative, second derivative, multivariate scattering correction, standard normal variable transformation, detrending, and Savitsky-Golay smoothing; Preferably, the multivariate correction algorithm is selected from one or more of the following: multivariate linear regression, principal component regression, and partial least squares.

5. The test method according to any one of claims 1-4, characterized in that, In step (3), when establishing the quantitative analysis model, the spectral data used is selected from one of the following methods: a) Use absorbance data across all wavelengths collected; b) Select absorbance data within one or more wavelength ranges; c) Select absorbance data at multiple discrete characteristic wavelength points; Preferably, when establishing a quantitative analysis model, spectral data in the wavelength range of 1460 nm to 1660 nm are excluded.

6. The test method according to any one of claims 1-5, characterized in that, Before, during, and after spectral acquisition, the sample chamber of the spectrometer is continuously or intermittently purged with dry gas.

7. The test method according to any one of claims 1-6, characterized in that, In step (4), after obtaining the contents of ammonium nitrate and all additives using a quantitative analysis model, the water content is calculated by the difference method.

8. A near-infrared spectroscopy analysis system for implementing the test method according to any one of claims 1-7, characterized in that, The system includes: Near-infrared spectrometer, used to collect near-infrared spectra of standard samples and samples to be tested; Temperature control device, used to adjust and stabilize the temperature of the sample during spectral acquisition; The computing processing unit is configured as follows: Store the standard spectral dataset and the known content of each component; A multivariate calibration algorithm is executed to establish the quantitative analysis model. The spectrum of the sample to be tested is received, and the content of each component is calculated and output using the quantitative analysis model.

9. A method for constructing a standard sample set for quantitative analysis of hydroxyammonium nitrate propellants, characterized in that, The construction method includes: Based on the chemical composition and technical specifications of the hydroxyl ammonium nitrate propellant, the design concentration ranges of hydroxyl ammonium nitrate, various additives, and water are determined. Using ammonium nitrate, additives, water, and / or mixtures containing two or more propellant components as raw materials, multiple standard samples covering the designed concentration range are prepared by precise weighing and mixing. Preferably, different amounts of nitric acid or nitric acid aqueous solution are precisely added to some or all of the prepared standard samples to simulate the random distribution of free nitric acid in the actual product, thus forming an extended standard sample set.

10. A non-transitory computer-readable storage medium storing a computer program, characterized in that, When executed by a processor, the program is configured to implement a test method for the content of nitrate hydroxyl ammonium propellant components as described in any one of claims 1-7.

Images